I can't say I've compiled a kernel in years--the defaults are almost always fine for me.
I/do/ see many crashes under Linux with ATI cards, if you use the ATI binary driver. I only partially blame Linux for this--it'd be nice if it was robust enough to survive ATI's mangling, however any time you allow unknown code to touch kernel memory, you could get a crash that you couldn't predict.
Apple is so freakish about control that they won't ship their OS with drivers that will do this. Linux doesn't have the luxury of that control.
I'm writing this on a 4 year old Dell laptop, so I don't know if you're implying that Dells are unreliable or what, but that hasn't been my experience. I do have a few thoughts on this, however....
I think that Macs tend to last longer because their owners are more protective of them. Mac users, at some level, like the look of having a Mac. They like the design of the Mac, and they probably treat them more carefully. I've seen so many Mac users with protective sleeves for their notebooks that it's crazy. PCs, seen as a commodity, and generally looking more like a tool than a work of art, are treated as such.
Anyway, I've seen Macs crash plenty of times. Less than Windows, sure, but more than Linux.
Also, pricing Apples and Dells is pretty tricky. You can usually get steep discounts on Dells, and you basically never get discounts with Apple.
Not the original poster, but I'll respond anyway.
I have ubuntu at work and printing to samba printing was as simple as adding a printer. I haven't looked at smb.conf on the machine. The poster was probably referring to creating shares on an Ubuntu system, rather than accessing shares on one. I haven't tried to share files, but exporting a printer to Samba did not work seamlessly at all, despite the fact that my printer is well-supported in Linux. I did it the old-fashioned way--editing smb.conf--and now my other Linux machines can print to it, but OS X machines cannot. Hard to say where the problem is, and I'm not particularly inclined to troubleshoot it further, but just getting the initial share going was a pain.
Firewall is installed on ubuntu, and enabled by default. The lack of open ports on a default install makes this less of an issue too. I don't know what firewall you get by default in Ubuntu--I haven't seen one installed by default in any of the installations I've made, much less one that's on by default. Maybe I missed something?
disk encryption. If I have a zip file with a password then I can click on it, enter the password and browse/edit files on it using a finder-like interface. That seems very like disk encryption to me. I suppose there is no flashing neon-light saying 'disk encryption' though... maybe the next version will highlight it more. That's file/directory encryption. He's talking about encrypting either the whole disk or your home folder, and automatically decrypting/mounting upon supplying the correct passphrase. E.g. Bitlocker.
You're right on VPN--it's dead simple in Ubuntu 7.04. Not so much in the earlier versions. Same with sound and monitors.
Can you replace the hard drive in the iMac without voiding the warranty? How about adding a sound card or extra NIC (without saturating your entire USB BUS?)
But the thing that this system runs is OS X, which other systems can't run. Let's be clear: there's no fundamental reason that OS X can't run on the other systems except because Apple says so. When you buy a Mac, you're buying what amounts to commodity PC hardware in a pretty wrapper. The difference in price between identical hardware at Dell goes to pay for OS X. So when I configure a Dell system that's $500 less than an equivalent Mac, Apple is telling me that OS X is worth $500 (a little more, really, since Dell includes Windows on most models, and you can "return" it for between $30 and $70, effectively dropping the cost even more.)
OS X is great. I love it a lot. But it's really hard to support this kind of lock-in.
If someone using Windows bitched about his dongle not working and he blamed the manufacturer for sell him junk, you would not be harping on him (I assert). It may be true for a lot of people, because Windows is such a dominant OS. I, however, acknowledge the difference.
Back when I still ran Windows, I had the displeasure of owning an ATI All-in-Wonder graphics card. Periodic reinstalls were a chore, because the drivers were quite unstable. Install them in the wrong order, and pieces of your card wouldn't work (video capture, perhaps, or syncing with audio.) Install them before certain codecs were installed in Windows and the driver seemingly failed to understand those codecs (i.e. you couldn't capture directly to them.) Reinstalling was an absurd chore--skip one step and something was almost guaranteed not to work. I had detailed notes on the whole process that I had to follow any time Windows became too unstable to work in.
The thing is, I still loved and used the card. When it worked (when you had installed everything correctly), it was fantastic. The video capture was much better than the previous video capture card I'd used. The quality was better. The software for capturing and playback (separate from the drivers) was really quite good. The card had a dongle which you could send video/audio through so that you weren't cluttering up the back of your PC with more wires. Everything about the card was fantastic--so much so that when I upgraded my computer, I bought another ATI All-in-Wonder. Having already worked out the kinks in the driver and installer, it only made sense, as the card itself was fantastic. But you can bet that I bitched about ATI--not for selling me junk, but for seemingly being unable to program their driver installers robustly.
I bitch about ATI today, too. In Windows (which is still sadly necessary for gamers), ATI cards don't give me any trouble. In Linux, they lack compositing, and they crash more often than Nvidia drivers. I blame ATI--again, not for "sellilng me crap", but for crappy drivers.
No, I'd still harp on someone who made that claim, regardless of their OS of choice.
I AM ARGUING THAT LINUX HAS REACHED (or is reach*ing*) A PLACE WITHIN THE MARKET THAT BRINGS WITH IT THAT SAME LEVEL OF EXPECTATION. OS X has about 5% of the market share right now. Linux has 1-2%. Linux is not anywhere near the point where it's reasonable to expect the same level of manufacturer support as Windows. Now Linux has a lot going for it right now (what with two major manufacturers now offering to preinstall it on their PCs), but it has a long way to go. We don't yet know whether it's really going to make it into the homes of people who have never heard of it, or what the backlash will be when they can't download their crappy animated cursors, or get to their banking website (many of which still require IE), etc. I think you're jumping the gun a lot, here, and I think that you have high expectations of manufacturers. I wouldn't expect much support, even if Linux doubles its market share. Why? Because OS X doesn't have the support of most manufacturers. We'd probably need to see Linux get in-and-around the 15-20% before we start seeing managers take notice, and probably 30% before the majority of them realize that they need to devote some resources to Linux support. There's just too much money involved in training/hiring people to do development on a completely new platform, supporting it as well as you support your other drivers (via phone or e-mail or whatever--you've got to train your helpdesk on it, which means training them not only on your hardware and software, but also on Linux if they aren't familiar with it), etc.
No kidding. Did anything in my comment imply that I do not understand this? Yes. I think it was the part where you said:
I just go straight to "vendor x makes crap hardware". I read your entire post the first time. You certainly made clear that you know what the problem is, right up until the last statement. My Nvidia analogy should have indicated that I understood that--i.e. flipping a switch magically made the hardware not crap anymore. The problem is that you seem to put forth information which may be a false conclusion based on poor analysis. You'll say "X makes crap hardware" when it's not necessarily the hardware that's at fault. You're spreading fear, uncertainty, and doubt about the manufacturer just because they don't support your favored OS. Broadcom might work perfectly well on Windows--but you'll claim that they make crap hardware because of its poor Linux support.
I really can't believe that you got modded Insightful.
Although hardware and driver design can be tightly coupled, the quality of one does not imply anything about the quality of the other. For example, I've always considered Nvidia to make pretty good hardware. For a really long time, they did not provide a Linux driver. By your logic, it would seem, "Nvidia made crap hardware." Right up until someone decided to flip a switch and loose their driver upon the Linux community. By your reasoning, flipping that switch suddenly made all that hardware suddenly not suck. Your arguments are inane and FUD.
The simple truth is that sometimes manufacturers don't support all operating systems. I can't get Windows 95 drivers for many new hardware devices--that doesn't mean that they make crap hardware just because it doesn't support my OS of choice. Don't like that I'm talking about out-of-date software? It doesn't support the newest release of OpenBSD or FreeBSD, either, despite the fact that there may be a Linux binary driver.
You're obviously welcome not to buy hardware that isn't going to work with your system, but don't disparage that hardware. Disparage the company for refusing to support your OS if you must. Disparage them for not having open specs, certainly. But it's not a hardware problem.
NO. Language isn't a one-to-one, and your so-called 'key' is open to anyone who wants to read a book on English to Chinese translations. A real key, one which is used in encryption, is intended to be kept a secret.
I didn't see that anywhere in the article, so I can't really comment on it. If that's the 'attack' they were complaining about, though, it's definitely pretty lame.
But the poster to whom I replied said:
technically speaking all data is "encrypted" .
That said, in response to your post, it's simply too difficult to describe the difference between using esoteric utilities which have the appearance of cloaking, and cloaking itself. The article actually describes ways to booby-trap your hard drive in case forensics software starts looking at it. That goes beyond cloaking--it's a pretty active attack on forensics methods. Finding such an attack, which by its nature must be targetted at specific software, gives a strong indication that someone was intentionally trying to thwart forensics, whereas using some odd compression software (or perhaps your own compression software) really doesn't.
You said you know the factors of every prime. Go ahead. Tell them to me.
Knowing the factors of every prime indicates that you know all of the primes. Knowing the factors of any given prime is what you're claiming to be able to do with the statement
name a large prime number and I'll tell you its factors
Nah, I know the factors of any given prime number, but if I knew the factors of every prime number, I'd have some mathematical theories to disprove./pedantic
Or perhaps that you speak Chinese and deal with lots of Chinese people... or that you're learning Chinese and are trying to improve your skill...or that you download lots of Chinese content....
Honestly, something in another language isn't going to make me think that someone is trying to hide something at all. It's not a logical conclusion.
There's a huge semantic difference between encrypting and encoding. All data is encoded, however encryption implies that there exists a secret which must be known in order to recover the encoded data.
Now you can get pretty fuzzy in talking about whether or not strange filesystems constitute enough of a secret for them to be called encryption, however encodings such as ASCII, Unicode, Huffman codes, etc. are not encryption by either the popular or the cryptographic definitions.
The drunk thing is way outside the control of the administrators. Testing the failover is something they can do, and if something doesn't work, they can fix it.
All of the applications I use on my Linux desktop are developed with highly portable (yep, cross-platform) toolkits such as GTK+ and Qt. Most run very well on many architectures and many kernels (Linux 2.4 and 2.6, *BSD including Darwin, Solaris, etc). You must be new here.
On Slashdot, it's not cross-platform if there's a single platform that it doesn't run on. See the various complaints about Flash not being "cross-platform" since it doesn't natively run on 64-bit Linux.
But seriously, most of the time when people say "cross-platform," they mean that it should run natively on Windows, Linux, and OS X. You can often force Linux applications to run on Windows if there's not a native port, but it's usually a pain in the ass.
Can't test that nvidia driver in a virtual machine either. That's exactly what I said. "Containerized OS" is the same thing as a virtual machine.
I do hope that this at one point becoms integral to the operating system, where individual applications can be encapsulated (like a sandbox), but then even suspended/resumed and migrated between physical machines. I think that's where we're heading (again.)
You don't get anything similar on Linux, and generally speaking, these alternates can't run proprietary OS.
We've had UML and chroot for quite a while in Linux, but it's equally limited. With virtualization, I can run Windows on my Linux box, which is (to me) where the real use is.
chroot jails tend to be restrictive. You can't access all your entries in/dev, or if you can, you've removed a lot of the protection afforded by the jail in the first place.
Virtualization (or containerization... how awful!) generally allows this. Want to play with your hard drive driver? No problem.
Of course, it fails when you actually/do/ want direct access to the hardware. Can't test that new Nvidia driver in a containerized OS.
Slashdot Mirror
I can't say I've compiled a kernel in years--the defaults are almost always fine for me.
/do/ see many crashes under Linux with ATI cards, if you use the ATI binary driver. I only partially blame Linux for this--it'd be nice if it was robust enough to survive ATI's mangling, however any time you allow unknown code to touch kernel memory, you could get a crash that you couldn't predict.
I
Apple is so freakish about control that they won't ship their OS with drivers that will do this. Linux doesn't have the luxury of that control.
I'm writing this on a 4 year old Dell laptop, so I don't know if you're implying that Dells are unreliable or what, but that hasn't been my experience. I do have a few thoughts on this, however....
I think that Macs tend to last longer because their owners are more protective of them. Mac users, at some level, like the look of having a Mac. They like the design of the Mac, and they probably treat them more carefully. I've seen so many Mac users with protective sleeves for their notebooks that it's crazy. PCs, seen as a commodity, and generally looking more like a tool than a work of art, are treated as such.
Anyway, I've seen Macs crash plenty of times. Less than Windows, sure, but more than Linux.
Also, pricing Apples and Dells is pretty tricky. You can usually get steep discounts on Dells, and you basically never get discounts with Apple.
You're right on VPN--it's dead simple in Ubuntu 7.04. Not so much in the earlier versions. Same with sound and monitors.
Can you replace the hard drive in the iMac without voiding the warranty? How about adding a sound card or extra NIC (without saturating your entire USB BUS?)
OS X is great. I love it a lot. But it's really hard to support this kind of lock-in.
SSH tunnels are probably not the best choice for this. Check out stunnel, though. It was virtually made for this kind of thing.
Of course, you have to send the logs over TCP, but in practice, that's not a big deal.
Back when I still ran Windows, I had the displeasure of owning an ATI All-in-Wonder graphics card. Periodic reinstalls were a chore, because the drivers were quite unstable. Install them in the wrong order, and pieces of your card wouldn't work (video capture, perhaps, or syncing with audio.) Install them before certain codecs were installed in Windows and the driver seemingly failed to understand those codecs (i.e. you couldn't capture directly to them.) Reinstalling was an absurd chore--skip one step and something was almost guaranteed not to work. I had detailed notes on the whole process that I had to follow any time Windows became too unstable to work in.
The thing is, I still loved and used the card. When it worked (when you had installed everything correctly), it was fantastic. The video capture was much better than the previous video capture card I'd used. The quality was better. The software for capturing and playback (separate from the drivers) was really quite good. The card had a dongle which you could send video/audio through so that you weren't cluttering up the back of your PC with more wires. Everything about the card was fantastic--so much so that when I upgraded my computer, I bought another ATI All-in-Wonder. Having already worked out the kinks in the driver and installer, it only made sense, as the card itself was fantastic. But you can bet that I bitched about ATI--not for selling me junk, but for seemingly being unable to program their driver installers robustly.
I bitch about ATI today, too. In Windows (which is still sadly necessary for gamers), ATI cards don't give me any trouble. In Linux, they lack compositing, and they crash more often than Nvidia drivers. I blame ATI--again, not for "sellilng me crap", but for crappy drivers.
No, I'd still harp on someone who made that claim, regardless of their OS of choice. I AM ARGUING THAT LINUX HAS REACHED (or is reach*ing*) A PLACE WITHIN THE MARKET THAT BRINGS WITH IT THAT SAME LEVEL OF EXPECTATION. OS X has about 5% of the market share right now. Linux has 1-2%. Linux is not anywhere near the point where it's reasonable to expect the same level of manufacturer support as Windows. Now Linux has a lot going for it right now (what with two major manufacturers now offering to preinstall it on their PCs), but it has a long way to go. We don't yet know whether it's really going to make it into the homes of people who have never heard of it, or what the backlash will be when they can't download their crappy animated cursors, or get to their banking website (many of which still require IE), etc. I think you're jumping the gun a lot, here, and I think that you have high expectations of manufacturers. I wouldn't expect much support, even if Linux doubles its market share. Why? Because OS X doesn't have the support of most manufacturers. We'd probably need to see Linux get in-and-around the 15-20% before we start seeing managers take notice, and probably 30% before the majority of them realize that they need to devote some resources to Linux support. There's just too much money involved in training/hiring people to do development on a completely new platform, supporting it as well as you support your other drivers (via phone or e-mail or whatever--you've got to train your helpdesk on it, which means training them not only on your hardware and software, but also on Linux if they aren't familiar with it), etc.
I really can't believe that you got modded Insightful.
Although hardware and driver design can be tightly coupled, the quality of one does not imply anything about the quality of the other. For example, I've always considered Nvidia to make pretty good hardware. For a really long time, they did not provide a Linux driver. By your logic, it would seem, "Nvidia made crap hardware." Right up until someone decided to flip a switch and loose their driver upon the Linux community. By your reasoning, flipping that switch suddenly made all that hardware suddenly not suck. Your arguments are inane and FUD.
The simple truth is that sometimes manufacturers don't support all operating systems. I can't get Windows 95 drivers for many new hardware devices--that doesn't mean that they make crap hardware just because it doesn't support my OS of choice. Don't like that I'm talking about out-of-date software? It doesn't support the newest release of OpenBSD or FreeBSD, either, despite the fact that there may be a Linux binary driver.
You're obviously welcome not to buy hardware that isn't going to work with your system, but don't disparage that hardware. Disparage the company for refusing to support your OS if you must. Disparage them for not having open specs, certainly. But it's not a hardware problem.
It depends upon how far they go. Don't forget Fair Use.
Interestingly, C is no longer a proper subset of C++.
Yeah, in this case, the article didn't seem to have much connection to the summary at all.
NO. Language isn't a one-to-one, and your so-called 'key' is open to anyone who wants to read a book on English to Chinese translations. A real key, one which is used in encryption, is intended to be kept a secret.
I didn't see that anywhere in the article, so I can't really comment on it. If that's the 'attack' they were complaining about, though, it's definitely pretty lame.
That said, in response to your post, it's simply too difficult to describe the difference between using esoteric utilities which have the appearance of cloaking, and cloaking itself. The article actually describes ways to booby-trap your hard drive in case forensics software starts looking at it. That goes beyond cloaking--it's a pretty active attack on forensics methods. Finding such an attack, which by its nature must be targetted at specific software, gives a strong indication that someone was intentionally trying to thwart forensics, whereas using some odd compression software (or perhaps your own compression software) really doesn't.
Knowing the factors of every prime indicates that you know all of the primes. Knowing the factors of any given prime is what you're claiming to be able to do with the statement name a large prime number and I'll tell you its factors
Nah, I know the factors of any given prime number, but if I knew the factors of every prime number, I'd have some mathematical theories to disprove. /pedantic
Or perhaps that you speak Chinese and deal with lots of Chinese people... or that you're learning Chinese and are trying to improve your skill...or that you download lots of Chinese content....
Honestly, something in another language isn't going to make me think that someone is trying to hide something at all. It's not a logical conclusion.
There's a huge semantic difference between encrypting and encoding. All data is encoded, however encryption implies that there exists a secret which must be known in order to recover the encoded data.
Now you can get pretty fuzzy in talking about whether or not strange filesystems constitute enough of a secret for them to be called encryption, however encodings such as ASCII, Unicode, Huffman codes, etc. are not encryption by either the popular or the cryptographic definitions.
Nope. They're not talking some random drunk off of the street, they're talking about a disgruntled employee.
The drunk thing is way outside the control of the administrators. Testing the failover is something they can do, and if something doesn't work, they can fix it.
On Slashdot, it's not cross-platform if there's a single platform that it doesn't run on. See the various complaints about Flash not being "cross-platform" since it doesn't natively run on 64-bit Linux.
But seriously, most of the time when people say "cross-platform," they mean that it should run natively on Windows, Linux, and OS X. You can often force Linux applications to run on Windows if there's not a native port, but it's usually a pain in the ass.
You don't get anything similar on Linux, and generally speaking, these alternates can't run proprietary OS.
We've had UML and chroot for quite a while in Linux, but it's equally limited. With virtualization, I can run Windows on my Linux box, which is (to me) where the real use is.
chroot jails tend to be restrictive. You can't access all your entries in /dev, or if you can, you've removed a lot of the protection afforded by the jail in the first place.
/do/ want direct access to the hardware. Can't test that new Nvidia driver in a containerized OS.
Virtualization (or containerization... how awful!) generally allows this. Want to play with your hard drive driver? No problem.
Of course, it fails when you actually