Re:Postscript is a complete language
on
PDF Virus Spotted
·
· Score: 2, Interesting
I've been thinking about this for a while (after playing with GILT).
Lack of I/O facilities means you couldn't create a postscript file that could replicate, but you could still potentially cause a bit of havoc. For example, create a postscript file that uses the random number generator to either print an amusing poster (99.9% of the time) or print several pages of dirty pictures (0.1% of the time). People will print the amusing document, send the file to all their friends, and eventually someone will get into trouble.
There are lots of people saying 'don't do this, it's not ethical'. The reasoning is clear: to fix a broken machine, you'd effectively have to 'hack' into it, which is bad; there's a chance that you could break something in the process, which is also bad. It's also probably illegal, although the odds are against you being found.
I'm not going to argue that it isn't unethical, but I do think it's still worth exploring the possibilities. Let's assume that CodeRedII isn't the last worm of its kind that we'll see. I find it fairly likely that this sort of attack will be more common, because of the possibilities it presents to hackers.
CodeRedI and II haven't brought the Internet to its knees yet, but the offer the potential for massive denial-of-service attacks that would have the power to do tremendous damage. If script-kiddies start taking advantage of this power, what should we do?
I wouldn't want to start trying to patch these systems myself, but I'd wouldn't complain about any vigilantes doing the job. Furthermore, it's possible that future CodeRed variants will not only attempt to infect other web servers, they'll also attempt to download new instructions from random web servers. If this started happening, I'd have no qualms about providing self-cleaning instructions on my own server.
The IIS weakness is found. The CodeRedII System goes on-line August 4th, 2001. Human decisions are removed from strategic hacking. CodeRedII begins to spread at a geometric rate. It becomes self-aware at 2:14AM, Eastern Time, August 29th.
One of the interesting side-effects of CodeRed that hasn't been discussed is this: anyone who wants to can now hack other machines with almost complete anonymity.
I've now got a huge list of IP addresses of badly administered machines with a known IIS backdoor. It's highly unlikely that anyone would notice my attempts to hack this machine over the background noise of CodeRed traffic flying around.
In a sense, CodeRed provides a smoke-screen for other hacking attempts, and a 'smoke-signal' to let hackers know where infected servers are.
No, your legal system doesn't have to make a call on whether or not Apple is a monopoly until they start using their position to squeeze other people out of the market. The Land Of The Free can hardly go around making it illegal to be a monopoly, but it does have enough of a sense of fair play to think about doing something about it when people abuse their monopoly position.
Apple haven't done this (err, enough). Microsoft haven't stopped doing this. That's the difference.
I think it's pointless trying to predict what will be the limiting factors for storage space requirements. You've decided that we don't need to store more information than the human eye can distinguish - and yet there are a number of reasons why this might not be the case: if you are storing data for future editing, you don't necessarily know what you are going to want to keep.
Video cameras used at sporting events can capture several hundred frames per second so that you can subsequently watch a smooth slow-motion replay. If you want to produce a 16-bit audio recording, you might use 24-bit samples during the editing process to avoid adding too much noise.
There is no hard limit to how much useful data you can store. You say that the detail for a huge 3D game will never need to go above a certain resolution, but I don't see how you can decide what the resolution will be. Should it be decided based on how much you can see if you're standing 1 metre away from an object? 10 centimetres? Looking at it through a microscope?
Ok, may be there is a theoretical limit when we model stuff at the atomic level, but I can't imagine that is what you are referring to when you say 'we are getting close enough to begin thinking about this realistically'.
...and I'm sure even then people will still find more ways of creating data.
'I could have ${danced} all night' is a song from My Fair Lady, the musical by Alan Lerner and Fred Loewe, based on Shaw's Pygmalion.
Julie Andrews played Eliza Doolittle in the original cast, 1956.
Julie Andrews also played Mary Poppins in the 1964 film of the same name, which features the atrocious-sounding song 'Supercalafragilisticexpialidocious'.
She also played Maria in The Sound Of Music (1965), but failed to sing either of the above songs in this role.
People who don't get the references to gannets wetting their nests might like to read this, which also mentions 'the Amazing Adventures of Captain Gladys Stoutpamphlet and her Intrepid Spaniel Stig amongst the Giant Pygmis of Beckles' (volume eight).
Using these facts, re-read the preceding conversation and see if it makes more sense now.
I don't mind a company keeping track of this. If I call up my bank and the guy said 'hi, how are the kids doing', I'd be - well, surprised, because I don't have any kids, but assuming I did, I'd feel pleased at the fact that they took the time out to record these things. And it doesn't matter that the person being pally with me is someone I've never spoken to before; I'm speaking to a representative of the bank, who is treating me as a valued customer.
It would get scary if all companies pooled this information though. If there were terabytes of data on everyone's preferences, home life, and conversation style, and any company that wanted to be pally could just buy into this database and instantly become your friend.
I like to think that in 100 years time, we will have computers that would scoff at this definition of intelligence. Why should machine 'intelligence' be hampered by the misfeatures of human intelligence?
Computers might some day ask themselves 'I wonder if biological organisms could ever be truly intelligent?' - except they'd be intelligent enough to realise that just as Linux can be ported to different processors, so intelligence (artificial or not) can run on wetware or hardware with equal ease.
I can't remember being very impressed by any of the arguments in Penrose's 'Emperor's New Mind'. May be I should read it again for a laugh. -- Andrem
I'm sure everyone noticed that Dr. Vosknocker in South Park: BL&U was played by Eric Idle. And don't forget John Cleese as 'The Bomb' and Terry Jones as the Parrot in Douglas Adams' computer game Starship Titanic. -- Andrem
Equally importantly, why should people have to feel afraid of being shouted down when they ask questions like yours? People should be able to have calm and rational debate over touchy issues. It's almost as if most people can't make a distinction between talking about doing something, and the act itself. -- Andrem
It's better than that: Nature summarise an article, then/. gives us a precis of the Nature article. According to the abstract of the original article, 'one can discover the outcome of [a] computation [...] without running the computer', and, of course, 'there are some limits on the information that can be obtained from them'.
Nature seems to have turned this into 'it should be possible to determine the outcome of a computation while the machine stays off', and not told us anything about what the limits are.
May be the Nature reporter didn't want to observe the original article for fear of collapsing its waveform.
It would be dead handy if they could extend their research to show that you don't even have to build the quantum computer in the first place. Hey, you probably didn't even need to know the results of the calculation, do you?
Normally 3D graphics work by projecting the scene onto a flat plane: draw lines from the player's eye to objects in the virtual world, through a plane in front of the eye. If this line hits a blue object, paint the corresponding point on the plane blue. Cut out a rectangular portion of this plane and stick it on the computer screen.
A true panorama view works by projecting the scene onto a cylinder centered on the player, then unfolding this cylinder, cutting out a rectangular portion and sticking it on the screen.
Judging by the comments about 'slices' in the readme file, I'd say that this hack works by approximating the cylindrical projection using a number of tall, thin rectangles. This isn't totally accurate, but I suspect most graphic cards don't support true cylindrical projections.
Other projections are possible: if you project the scene onto a sphere, you can unfold part of the sphere onto the rectangular computer screen using the same techniques for drawing maps of the earth on flat paper...
Of course, if your computer screen isn't a flat rectangle then you can really have fun... some people have mentioned using PanQuake to render onto hemispherical screen: if you want to do this, you are best of working out the maths behind the whole projection (from 3d screen to 2d graphics card memory to projector back to 3d screen) and hacking the rendering engine to do the projection correctly. And then make sure you are sitting in exactly the right place to get the best view... -- Andrem
Re:Playing Quake like this causing health problems
on
PanQuake
·
· Score: 1
I've found that a really good way of inducing motion sickness in FPS games is to play with a cordless mouse. These seem to transmit potion updates at a relatively low frequency - so if the game is refreshing the screen 60 times a second, but the mouse position is only being updated 25 times a second, and you are using the mouse to look and the keyboard to run, you will get continuous forward motion, with intermittent rotational motion. As a result, I can't play these sorts of games with a cordless mouse for very long. -- Andrem
Everything I've read about XP suggests that it is being used by people who are developing bespoke applications, where each program you develop IS catering to one customer's needs.
I used to work in a company that did this sort of thing. There was a lot of R&D effort expended in trying to create a generic appliation base that could speed up project development, but mostly this came from making libraries out of good bits of code from existing projects, rather than designing anything useful in advance. -- Andrem
Yes we can say software crisis. Take a bunch of requirements, spend months designing a wonderful class library that can do everything that is required, set the code-monkeys loose writing it, and then find that it doesn't do what the customer wanted.
I don't think XP is a magic bullet that will make the software crisis go away - but on the other hand, sticking rigidly to Principles of Good Software Design that have been tried and tested for decades and have yet to consistently deliver quality products on time isn't going to do any good either.
Oh, and claiming that the central tenet of XP is that you don't bother designing anything seems to be a bit wide of the mark. If I had to put only one aspect of XP into practise, it would be writing unit tests before writing each unit of code. I've tried this a few times, and it has paid dividends - and it isn't possible to do this without having some kind of design in mind before you start. -- Andrem
Ooh - that brings back memories of frantically switching cables over after the tape had finished loading so that you could get the full glory of the dramatic words 'ROBOT ATTACK'.
Dancing Demon was great! -- Andrem
What's the problem? You buy some PCs from your friendly local vendor for your Linux-only company. You encourage the vendor to report you to Microsoft, the vendor gets free ice-cream from Microsoft for being good and reporting the full frontal nudity of the purchase, and Microsoft can do didly squat about it. -- Andrem
'the dangers to health increase exponentially as the size of the screen'.
I can only assume we're talking about the angular size of the screen: a 14" screen will produce exactly the same image on your retina as a 28" screen twice the distance away.
'The researchers attributed this change in brain activity to high frequency components in the rapidly changing images at the side of the screen'.
So if you're completely surrounded by screens, there are no edge effects to worry about? -- Andrem
We are now automatically hiring anyone who has released any music recordings under the GNU Pulic Music License. If any suitable candidates are found, please forward their details to recruitment@burgerking.com.
Note, you may want to consider laugh and their gullibility first first.
Depending on the size of the allocation blocks and the file in question, you could potentially make savings using an ADFS-like file-system, and you wouldn't even have to use the trick of removing every copy of a certain byte from the file. The design of the file-system allows small files to be squeezed into spaces that large files wouldn't fit in: large files have to use space allocated in blocks (which have a minimum size - for example, 30 sectors).
Several small files can share one block - so you could fit one directory catalog and several small files into a single block, but a large file would have to have a whole block to itself.
So, if the minimum block size were 60k, and a directory catalog took 2k, then a directory+70k file would require 180k, whereas directory+20k file+50k file would require 120k.
Of course, no-one sensible would claim that this consitutes compression of the original file: I've come up with a way of storing a 70k file in only 120k! However, it is an improvement over the 180k you'd normally get, and on occasion it could be useful to know about this technique (like when ADFS reported 0 bytes free on a floppy disk, you could normally guarantee that there would be room for a few more 512 byte files...)
Slashdot Mirror
> the adobe one
...there's a hole in my bucket
Lack of I/O facilities means you couldn't create a postscript file that could replicate, but you could still potentially cause a bit of havoc. For example, create a postscript file that uses the random number generator to either print an amusing poster (99.9% of the time) or print several pages of dirty pictures (0.1% of the time). People will print the amusing document, send the file to all their friends, and eventually someone will get into trouble.
There are lots of people saying 'don't do this, it's not ethical'. The reasoning is clear: to fix a broken machine, you'd effectively have to 'hack' into it, which is bad; there's a chance that you could break something in the process, which is also bad. It's also probably illegal, although the odds are against you being found.
I'm not going to argue that it isn't unethical, but I do think it's still worth exploring the possibilities. Let's assume that CodeRedII isn't the last worm of its kind that we'll see. I find it fairly likely that this sort of attack will be more common, because of the possibilities it presents to hackers.
CodeRedI and II haven't brought the Internet to its knees yet, but the offer the potential for massive denial-of-service attacks that would have the power to do tremendous damage. If script-kiddies start taking advantage of this power, what should we do?
I wouldn't want to start trying to patch these systems myself, but I'd wouldn't complain about any vigilantes doing the job. Furthermore, it's possible that future CodeRed variants will not only attempt to infect other web servers, they'll also attempt to download new instructions from random web servers. If this started happening, I'd have no qualms about providing self-cleaning instructions on my own server.
May be the latest versions of PGP include the time between key presses as part of the passphrase?
The IIS weakness is found. The CodeRedII System goes on-line August 4th, 2001. Human decisions are removed from strategic hacking. CodeRedII begins to spread at a geometric rate. It becomes self-aware at 2:14AM, Eastern Time, August 29th.
One of the interesting side-effects of CodeRed that hasn't been discussed is this: anyone who wants to can now hack other machines with almost complete anonymity.
I've now got a huge list of IP addresses of badly administered machines with a known IIS backdoor. It's highly unlikely that anyone would notice my attempts to hack this machine over the background noise of CodeRed traffic flying around.
In a sense, CodeRed provides a smoke-screen for other hacking attempts, and a 'smoke-signal' to let hackers know where infected servers are.
No, your legal system doesn't have to make a call on whether or not Apple is a monopoly until they start using their position to squeeze other people out of the market. The Land Of The Free can hardly go around making it illegal to be a monopoly, but it does have enough of a sense of fair play to think about doing something about it when people abuse their monopoly position.
Apple haven't done this (err, enough). Microsoft haven't stopped doing this. That's the difference.
I think it's pointless trying to predict what will be the limiting factors for storage space requirements. You've decided that we don't need to store more information than the human eye can distinguish - and yet there are a number of reasons why this might not be the case: if you are storing data for future editing, you don't necessarily know what you are going to want to keep.
Video cameras used at sporting events can capture several hundred frames per second so that you can subsequently watch a smooth slow-motion replay. If you want to produce a 16-bit audio recording, you might use 24-bit samples during the editing process to avoid adding too much noise.
There is no hard limit to how much useful data you can store. You say that the detail for a huge 3D game will never need to go above a certain resolution, but I don't see how you can decide what the resolution will be. Should it be decided based on how much you can see if you're standing 1 metre away from an object? 10 centimetres? Looking at it through a microscope?
Ok, may be there is a theoretical limit when we model stuff at the atomic level, but I can't imagine that is what you are referring to when you say 'we are getting close enough to begin thinking about this realistically'.
...and I'm sure even then people will still find more ways of creating data.
'I could have ${danced} all night' is a song from My Fair Lady, the musical by Alan Lerner and Fred Loewe, based on Shaw's Pygmalion.
Julie Andrews played Eliza Doolittle in the original cast, 1956.
Julie Andrews also played Mary Poppins in the 1964 film of the same name, which features the atrocious-sounding song 'Supercalafragilisticexpialidocious'.
She also played Maria in The Sound Of Music (1965), but failed to sing either of the above songs in this role.
People who don't get the references to gannets wetting their nests might like to read this, which also mentions 'the Amazing Adventures of Captain Gladys Stoutpamphlet and her Intrepid Spaniel Stig amongst the Giant Pygmis of Beckles' (volume eight).
Using these facts, re-read the preceding conversation and see if it makes more sense now.
May be NASA could increase their publicity by sending allowing the public to vote out on crew member every couple of weeks
I don't mind a company keeping track of this. If I call up my bank and the guy said 'hi, how are the kids doing', I'd be - well, surprised, because I don't have any kids, but assuming I did, I'd feel pleased at the fact that they took the time out to record these things. And it doesn't matter that the person being pally with me is someone I've never spoken to before; I'm speaking to a representative of the bank, who is treating me as a valued customer.
It would get scary if all companies pooled this information though. If there were terabytes of data on everyone's preferences, home life, and conversation style, and any company that wanted to be pally could just buy into this database and instantly become your friend.
I like to think that in 100 years time, we will have computers that would scoff at this definition of intelligence. Why should machine 'intelligence' be hampered by the misfeatures of human intelligence?
Computers might some day ask themselves 'I wonder if biological organisms could ever be truly intelligent?' - except they'd be intelligent enough to realise that just as Linux can be ported to different processors, so intelligence (artificial or not) can run on wetware or hardware with equal ease.
I can't remember being very impressed by any of the arguments in Penrose's 'Emperor's New Mind'. May be I should read it again for a laugh.
-- Andrem
I'm sure everyone noticed that Dr. Vosknocker in South Park: BL&U was played by Eric Idle. And don't forget John Cleese as 'The Bomb' and Terry Jones as the Parrot in Douglas Adams' computer game Starship Titanic.
-- Andrem
In other words, you won't?
-- Andrem
Equally importantly, why should people have to feel afraid of being shouted down when they ask questions like yours? People should be able to have calm and rational debate over touchy issues. It's almost as if most people can't make a distinction between talking about doing something, and the act itself.
-- Andrem
It's better than that: Nature summarise an article, then /. gives us a precis of the Nature article. According to the abstract of the original article, 'one can discover the outcome of [a] computation [...] without running the computer', and, of course, 'there are some limits on the information that can be obtained from them'.
Nature seems to have turned this into 'it should be possible to determine the outcome of a computation while the machine stays off', and not told us anything about what the limits are.
May be the Nature reporter didn't want to observe the original article for fear of collapsing its waveform.
It would be dead handy if they could extend their research to show that you don't even have to build the quantum computer in the first place. Hey, you probably didn't even need to know the results of the calculation, do you?
-- Andrem
Normally 3D graphics work by projecting the scene onto a flat plane: draw lines from the player's eye to objects in the virtual world, through a plane in front of the eye. If this line hits a blue object, paint the corresponding point on the plane blue. Cut out a rectangular portion of this plane and stick it on the computer screen.
A true panorama view works by projecting the scene onto a cylinder centered on the player, then unfolding this cylinder, cutting out a rectangular portion and sticking it on the screen.
Judging by the comments about 'slices' in the readme file, I'd say that this hack works by approximating the cylindrical projection using a number of tall, thin rectangles. This isn't totally accurate, but I suspect most graphic cards don't support true cylindrical projections.
Other projections are possible: if you project the scene onto a sphere, you can unfold part of the sphere onto the rectangular computer screen using the same techniques for drawing maps of the earth on flat paper...
Of course, if your computer screen isn't a flat rectangle then you can really have fun... some people have mentioned using PanQuake to render onto hemispherical screen: if you want to do this, you are best of working out the maths behind the whole projection (from 3d screen to 2d graphics card memory to projector back to 3d screen) and hacking the rendering engine to do the projection correctly. And then make sure you are sitting in exactly the right place to get the best view...
-- Andrem
I've found that a really good way of inducing motion sickness in FPS games is to play with a cordless mouse. These seem to transmit potion updates at a relatively low frequency - so if the game is refreshing the screen 60 times a second, but the mouse position is only being updated 25 times a second, and you are using the mouse to look and the keyboard to run, you will get continuous forward motion, with intermittent rotational motion. As a result, I can't play these sorts of games with a cordless mouse for very long.
-- Andrem
Everything I've read about XP suggests that it is being used by people who are developing bespoke applications, where each program you develop IS catering to one customer's needs.
I used to work in a company that did this sort of thing. There was a lot of R&D effort expended in trying to create a generic appliation base that could speed up project development, but mostly this came from making libraries out of good bits of code from existing projects, rather than designing anything useful in advance.
-- Andrem
Yes we can say software crisis. Take a bunch of requirements, spend months designing a wonderful class library that can do everything that is required, set the code-monkeys loose writing it, and then find that it doesn't do what the customer wanted. I don't think XP is a magic bullet that will make the software crisis go away - but on the other hand, sticking rigidly to Principles of Good Software Design that have been tried and tested for decades and have yet to consistently deliver quality products on time isn't going to do any good either. Oh, and claiming that the central tenet of XP is that you don't bother designing anything seems to be a bit wide of the mark. If I had to put only one aspect of XP into practise, it would be writing unit tests before writing each unit of code. I've tried this a few times, and it has paid dividends - and it isn't possible to do this without having some kind of design in mind before you start.
-- Andrem
Ooh - that brings back memories of frantically switching cables over after the tape had finished loading so that you could get the full glory of the dramatic words 'ROBOT ATTACK'. Dancing Demon was great!
-- Andrem
What's the problem? You buy some PCs from your friendly local vendor for your Linux-only company. You encourage the vendor to report you to Microsoft, the vendor gets free ice-cream from Microsoft for being good and reporting the full frontal nudity of the purchase, and Microsoft can do didly squat about it.
-- Andrem
'the dangers to health increase exponentially as the size of the screen'.
I can only assume we're talking about the angular size of the screen: a 14" screen will produce exactly the same image on your retina as a 28" screen twice the distance away.
'The researchers attributed this change in brain activity to high frequency components in the rapidly changing images at the side of the screen'.
So if you're completely surrounded by screens, there are no edge effects to worry about?
-- Andrem
Burger King, 2001
FAO all recruitment agencies,
We are now automatically hiring anyone who has released any music recordings under the GNU Pulic Music License. If any suitable candidates are found, please forward their details to recruitment@burgerking.com.
Note, you may want to consider laugh and their gullibility first first.
-- Andrem
Depending on the size of the allocation blocks and the file in question, you could potentially make savings using an ADFS-like file-system, and you wouldn't even have to use the trick of removing every copy of a certain byte from the file. The design of the file-system allows small files to be squeezed into spaces that large files wouldn't fit in: large files have to use space allocated in blocks (which have a minimum size - for example, 30 sectors).
Several small files can share one block - so you could fit one directory catalog and several small files into a single block, but a large file would have to have a whole block to itself.
So, if the minimum block size were 60k, and a directory catalog took 2k, then a directory+70k file would require 180k, whereas directory+20k file+50k file would require 120k.
Of course, no-one sensible would claim that this consitutes compression of the original file: I've come up with a way of storing a 70k file in only 120k! However, it is an improvement over the 180k you'd normally get, and on occasion it could be useful to know about this technique (like when ADFS reported 0 bytes free on a floppy disk, you could normally guarantee that there would be room for a few more 512 byte files...)
-- Andrem