Beat the Burglar might only have targeted volunteers, but the more recent The Real Hustle didn't. (In one episode they went and fraudulently tricked a locksmith into opening someone else's house, then went in and installed secret cameras and stole things from it. Presumably according to BBC reasoning that's OK because they gave the things back and got permission to show the footage.)
Given that the BBC has a program which, amongst other things, steals things from people then gives them back again (and then gets permission to show the resulting film on TV), I suspect if they were going to get in trouble for this sort of thing they would have done long ago.
Once the BBC had finished with their botnet, they changed the desktop background of all the infected computers to tell people what had happened and link them to this webpage, which contains some information on how to secure Windows. Then, they uninstalled the botnet software.
And always remember to check the return value of the close, under any OS. It's entirely possible for an attempt to a close a file to fail if the disk is full. (And yes, there are ways to recover from this; save on a different filesystem, don't delete the backup/original, keep the file in memory so the user can free up space so you can save it.) Although IIRC this wasn't guaranteed, in my experience DOS tended to do nothing at all to a file opened for write until the file was closed.
KDE 4.2 is in the Ubuntu 8.10 "proposed" repositories, mostly (the packaging for Kate is broken at the moment). Running Ubuntu-proposed is probably a bad idea for anything but a personal or experimental machine, though, as it breaks every now and then. As you don't like unstable distros, it's almost certainly a bad idea for you, but there might be other people around here who want to try it out.
Actually, the "standard" defragmentation utility on Linux is tar, although it rarely needs to be used. If for some reason you want a defragmented filesystem (say you want to send the entire filesystem to someone for some reason and want the resulting file size to be as small as possible when compressed), you can do it by tarballing the system up and untarring it again (although that isn't an in-place defragment, so takes twice as long and requires the system to be no more than half full). A separate defragmenter is therefore an optimisation for a rarely-needed task that can be performed by other means, so no wonder they're rare to find on Linux.
Well, as a student engineer who switched to Linux during their course (as soon as I got my first computer of my own), I can give a quick summary of my experiences. I mostly continued using the same software I was before (any excuse to use a UNIX-like shell with standard UNIX utilities before that, though, so I was rather unusual). Most of the specialist software engineers use is installed on computers for the purpose anyway that I don't own, therefore still on Windows (having reflashing hardware attached to the computer is very useful when using compilers to generate the code to reflash, for instance), so that's been less of a problem that might be expected. The software that I do need to run here has so far worked fine in WINE; it has an extra-high tendency to work, due at least partly to the fact that it tends to be old and targeted at older versions of Windows anyway. (Not to mention, that many of the most popular engineering packages of all have Linux versions nowadays, although the vast majority by proportion don't.) (Incidentally, I install Windows software under its own username, to stop it messing with my main account; there are "security" reasons to not run as root and keep proper user encapsulation on a single-user account!) Nowadays I get almost all software via the package manager, which is insanely useful (when I have to switch to a new programming language I can normally get a development environment for it running in less than 5 minutes, for instance.) The main advantage over the Windows computers actually seems to be speed; this computer tends to lock up a lot less often than the Windows ones do (I've spend over 10 minutes staring at a frozen screen on a Windows computer before, which is admittedly unusual and excessive, but because it was locked down I had no chance to look into what was causing the problem.)
I think I adjusted to the change unusually well, but then I'm pretty unusual myself (having preferred UNIX to Windows all along when possible), and managed to avoid most of the common beginner's pitfalls. I suspect a different test case would be more useful.
And I chose the platform not for the availability of software, but for the reliability of the software once it's installed, and also for cost (as a student, I don't really want to spend the money on a Mac, nor even pay the Microsoft Tax if I can help it...)
At least one beginner user has. I tend to hang around #ubuntu during releases, and have helped users to sort out messes that have been made when the upgrades have caused package management related problems. (Beginning users know how to troubleshoot a lot of things; a pretty reliable method is to find someone who knows, and ask them. A less reliable method is to ask on IRC, but that often works too.)
Incidentally, dpkg -a --force-depends, followed by apt-get dist-upgrade, followed by using synaptic to fix broken packages, is a relatively good method for sorting out even really weird package management borkage on DEB-based systems. I wouldn't expect beginning users, or even most power users, to know that, though, or even have to know that. (The last time I used it was when I had to do a hard reboot in the middle of upgrading Ubuntu Hardy to Ubuntu Intrepid due to hardware problems. That's pretty likely to screw up your packaging system. On the other hand, I doubt there's any way, other than reinstalling, to recover from an upgrade from XP to Vista where you lost power halfway through...)
Ubuntu (in particular, I don't know about whether other Linux distributions do this) also has an even easier to use cut-down version of Synaptic called Applications | Add/Remove... No good for installing most command-line applications, but people who are scared of apt probably don't want those anyway (and can use Synaptic if they do).
Those passwords are the items of real value now, and they're in unprivileged processes within the reach of unprivileged malware.
I agree with you about the passwords; however, the botnet thing, less so. No, you don't need to be an administrator to be part of a botnet. However, good luck hiding from antivirus programs or the like without administrator (or higher) access. Non-administrator malware can do all it needs to do (and for stealing passwords, that can just be done once and your job's done); but hiding to avoid being removed and shrinking your botnet size?
Windows NT/XP has a perfectly good security model, if only people would use it. In some ways it's more sophisticated than Linux's: For instance, file permissions are more fine-grained on NT.
When that breaks, it can break spectacularly. The Windows XP systems here are locked down. On one memorable occasion, an admin was trying to install new software, and the install went really wrong for no obvious reason. It turned out the installer was creating files under a temporary filename then renaming them to the correct filename (sort of like.part downloads from Firefox), and for some reason the directory it had created had permissions to create files but not to rename files in it! I agree with that sort of thing in principle, but something seems to have gone rather wrong with the Windows implementation; in particular, the defaults for it normally come out completely wrong.
I'm one of the few people who (when they use Windows Vista, which is rare, although it's my favourite version of Windows so far precisely because of UAC) leaves UAC on, and makes sure that I know why a prompt came up before clicking allow. (By default, I click cancel unless I was expecting the prompt, and nothing seems to have gone wrong yet.) UAC is, as mentioned several times above, a good security model if done properly, and remarkably similar to the one generally used by Linux-based systems and used by Mac OS X. The problems are lack of information about why it's there, and the huge numbers of programs not designed for it. (Also the lack of setgid on Windows, I don't know if it's possible or not, but it's certainly not commonly used. That's a useful alternative to dangerous setuid or even more dangerous elevate-everything-by-not-reading-prompts. What do games on Windows do for things like high score tables, for instance? Localise per-user?)
Some languages also allow direct manipulation of the state of a closure (e.g. some Smalltalks allow explicit re-binding of variables in closures).
Well, yes, but Smalltalk allows anything no matter how ridiculous or paradigm-breaking it is. Try running "true become: false" some time, and watch your Smalltalk interpreter blow up. (Yes, some versions of Smalltalk accept that.) So saying that Smalltalk accepts something doesn't mean that the paradigm, in general, accepts it.
Nowadays fish and chips in England are sold on newsprint-quality paper, but not actually printed; they're packaged in the low-quality paper used to make newspapers, not newspapers themselves. So probably pretty easily, given that they already do without newspapers.
A warranty disclaimer is different from an EULA, although they're often bundled together. I think they're valid even if you don't specifically agree to them. However, they have to be prominent; that's apparently why they're always written in all-caps. (The GPL has a warranty disclaimer in too; and it's in all-caps too.)
IANAL, by the way.
I've learnt the trick of filing bugs with Debian instead, where they're applicable (i.e. not something Ubuntu-specific). Debian devs tend to be more responsive to bug reports, and the changes will make their way into Ubuntu from there. On the other hand, if it's an Ubuntu-specific problem (and it often is), this won't help.
Not that you'd ever want to do it like this anyway unless you were patching it yourself (you'd just get the binaries from the package manager) but:
apt-get source virtualbox-ose
cd virtualbox-ose-2.0.4-dfsg/
sudo apt-get build-dep virtualbox-ose
make -f debian/rules build
If anyone still believes that dependency hell is a problem in modern Linux distributions, I advise them to look at the third line of what I did above and be disillusioned; nothing about what I did apart from the filenames depends in any way on VirtualBox, I could have used any other package instead. (This particular technique only works for.DEBs, but both RPMs and portage are equally capable of solving the problem in their own ways, and I suspect most other Linux package managers can too.)
If the printer is cheaper than the ink, then why not just buy a new printer when it's time to change the ink? A lot of printer sales and no ink sales will send quite a strong message to the manufacturer...
Umm... that second link just gives me an error message, saying "Enable javascript to use LMGTFY." Why did you link to a website that needs JavaScript when you have several perfectly good search engines to turn to?
You mean just 3dB. Decibels measure ratios; a 3dB difference is approximately equal to a 2:1 ratio, or a doubling, which is presumably what you meant. When decibels are used with another unit, such as your dB$, the extra unit refers to the amount that "0 dB", or a ratio of 1:1 refers to. So 3dB = double (approximately, not exactly), 3dB$ = double one dollar = $2, which is probably not what you meant. (To put it another way, 33dB$/30dB$ = 33dB-30dB = 3dB; 33dB$-30dB$ = approx $2000-$1000 = approx $1000 = approx 30dB$.) Logarithmic units can be confusing, but given that you seem to have been trying to use them to show off, I would have expected you to get them right...
Slashdot Mirror
Beat the Burglar might only have targeted volunteers, but the more recent The Real Hustle didn't. (In one episode they went and fraudulently tricked a locksmith into opening someone else's house, then went in and installed secret cameras and stole things from it. Presumably according to BBC reasoning that's OK because they gave the things back and got permission to show the footage.)
Given that the BBC has a program which, amongst other things, steals things from people then gives them back again (and then gets permission to show the resulting film on TV), I suspect if they were going to get in trouble for this sort of thing they would have done long ago.
Once the BBC had finished with their botnet, they changed the desktop background of all the infected computers to tell people what had happened and link them to this webpage, which contains some information on how to secure Windows. Then, they uninstalled the botnet software.
And always remember to check the return value of the close, under any OS. It's entirely possible for an attempt to a close a file to fail if the disk is full. (And yes, there are ways to recover from this; save on a different filesystem, don't delete the backup/original, keep the file in memory so the user can free up space so you can save it.) Although IIRC this wasn't guaranteed, in my experience DOS tended to do nothing at all to a file opened for write until the file was closed.
KDE 4.2 is in the Ubuntu 8.10 "proposed" repositories, mostly (the packaging for Kate is broken at the moment). Running Ubuntu-proposed is probably a bad idea for anything but a personal or experimental machine, though, as it breaks every now and then. As you don't like unstable distros, it's almost certainly a bad idea for you, but there might be other people around here who want to try it out.
Actually, the "standard" defragmentation utility on Linux is tar, although it rarely needs to be used. If for some reason you want a defragmented filesystem (say you want to send the entire filesystem to someone for some reason and want the resulting file size to be as small as possible when compressed), you can do it by tarballing the system up and untarring it again (although that isn't an in-place defragment, so takes twice as long and requires the system to be no more than half full). A separate defragmenter is therefore an optimisation for a rarely-needed task that can be performed by other means, so no wonder they're rare to find on Linux.
Google are selling footguns here, rather than using them. Whether other people choose to buy them is entirely up to them.
And I chose the platform not for the availability of software, but for the reliability of the software once it's installed, and also for cost (as a student, I don't really want to spend the money on a Mac, nor even pay the Microsoft Tax if I can help it...)
At least one beginner user has. I tend to hang around #ubuntu during releases, and have helped users to sort out messes that have been made when the upgrades have caused package management related problems. (Beginning users know how to troubleshoot a lot of things; a pretty reliable method is to find someone who knows, and ask them. A less reliable method is to ask on IRC, but that often works too.) Incidentally, dpkg -a --force-depends, followed by apt-get dist-upgrade, followed by using synaptic to fix broken packages, is a relatively good method for sorting out even really weird package management borkage on DEB-based systems. I wouldn't expect beginning users, or even most power users, to know that, though, or even have to know that. (The last time I used it was when I had to do a hard reboot in the middle of upgrading Ubuntu Hardy to Ubuntu Intrepid due to hardware problems. That's pretty likely to screw up your packaging system. On the other hand, I doubt there's any way, other than reinstalling, to recover from an upgrade from XP to Vista where you lost power halfway through...)
Ubuntu (in particular, I don't know about whether other Linux distributions do this) also has an even easier to use cut-down version of Synaptic called Applications | Add/Remove... No good for installing most command-line applications, but people who are scared of apt probably don't want those anyway (and can use Synaptic if they do).
Those passwords are the items of real value now, and they're in unprivileged processes within the reach of unprivileged malware.
I agree with you about the passwords; however, the botnet thing, less so. No, you don't need to be an administrator to be part of a botnet. However, good luck hiding from antivirus programs or the like without administrator (or higher) access. Non-administrator malware can do all it needs to do (and for stealing passwords, that can just be done once and your job's done); but hiding to avoid being removed and shrinking your botnet size?
Windows NT/XP has a perfectly good security model, if only people would use it. In some ways it's more sophisticated than Linux's: For instance, file permissions are more fine-grained on NT.
When that breaks, it can break spectacularly. The Windows XP systems here are locked down. On one memorable occasion, an admin was trying to install new software, and the install went really wrong for no obvious reason. It turned out the installer was creating files under a temporary filename then renaming them to the correct filename (sort of like .part downloads from Firefox), and for some reason the directory it had created had permissions to create files but not to rename files in it! I agree with that sort of thing in principle, but something seems to have gone rather wrong with the Windows implementation; in particular, the defaults for it normally come out completely wrong.
I'm one of the few people who (when they use Windows Vista, which is rare, although it's my favourite version of Windows so far precisely because of UAC) leaves UAC on, and makes sure that I know why a prompt came up before clicking allow. (By default, I click cancel unless I was expecting the prompt, and nothing seems to have gone wrong yet.) UAC is, as mentioned several times above, a good security model if done properly, and remarkably similar to the one generally used by Linux-based systems and used by Mac OS X. The problems are lack of information about why it's there, and the huge numbers of programs not designed for it. (Also the lack of setgid on Windows, I don't know if it's possible or not, but it's certainly not commonly used. That's a useful alternative to dangerous setuid or even more dangerous elevate-everything-by-not-reading-prompts. What do games on Windows do for things like high score tables, for instance? Localise per-user?)
Nah, it should have been Cancel/Allow. Then they could rely on employees clicking on Allow without actually reading the message...
Some languages also allow direct manipulation of the state of a closure (e.g. some Smalltalks allow explicit re-binding of variables in closures).
Well, yes, but Smalltalk allows anything no matter how ridiculous or paradigm-breaking it is. Try running "true become: false" some time, and watch your Smalltalk interpreter blow up. (Yes, some versions of Smalltalk accept that.) So saying that Smalltalk accepts something doesn't mean that the paradigm, in general, accepts it.
Nowadays fish and chips in England are sold on newsprint-quality paper, but not actually printed; they're packaged in the low-quality paper used to make newspapers, not newspapers themselves. So probably pretty easily, given that they already do without newspapers.
A warranty disclaimer is different from an EULA, although they're often bundled together. I think they're valid even if you don't specifically agree to them. However, they have to be prominent; that's apparently why they're always written in all-caps. (The GPL has a warranty disclaimer in too; and it's in all-caps too.) IANAL, by the way.
Better security if you load IE by accident. (Some programs default to loading IE rather than Firefox...)
I've learnt the trick of filing bugs with Debian instead, where they're applicable (i.e. not something Ubuntu-specific). Debian devs tend to be more responsive to bug reports, and the changes will make their way into Ubuntu from there. On the other hand, if it's an Ubuntu-specific problem (and it often is), this won't help.
Not that you'd ever want to do it like this anyway unless you were patching it yourself (you'd just get the binaries from the package manager) but:
If anyone still believes that dependency hell is a problem in modern Linux distributions, I advise them to look at the third line of what I did above and be disillusioned; nothing about what I did apart from the filenames depends in any way on VirtualBox, I could have used any other package instead. (This particular technique only works for .DEBs, but both RPMs and portage are equally capable of solving the problem in their own ways, and I suspect most other Linux package managers can too.)
If the printer is cheaper than the ink, then why not just buy a new printer when it's time to change the ink? A lot of printer sales and no ink sales will send quite a strong message to the manufacturer...
Umm... that second link just gives me an error message, saying "Enable javascript to use LMGTFY." Why did you link to a website that needs JavaScript when you have several perfectly good search engines to turn to?
Although 0x50000000 is also a Friday 13th, in 2012.
3dB$
You mean just 3dB. Decibels measure ratios; a 3dB difference is approximately equal to a 2:1 ratio, or a doubling, which is presumably what you meant. When decibels are used with another unit, such as your dB$, the extra unit refers to the amount that "0 dB", or a ratio of 1:1 refers to. So 3dB = double (approximately, not exactly), 3dB$ = double one dollar = $2, which is probably not what you meant. (To put it another way, 33dB$/30dB$ = 33dB-30dB = 3dB; 33dB$-30dB$ = approx $2000-$1000 = approx $1000 = approx 30dB$.) Logarithmic units can be confusing, but given that you seem to have been trying to use them to show off, I would have expected you to get them right...
name me one case of a trojan being detected via open source.
http://www.freedom-to-tinker.com/blog/felten/linux-backdoor-attempt-thwarted (admittedly, that's giving a URL not naming, but I think it was defeated so quickly it wasn't even given a name).
Now, name me one case of a trojan that wasn't created because the source code was closed.