If a publisher of a non-free computer operating system has announced that it declines to issue more security patches for that version, and you have discovered a remote vulnerability in that operating system version, it's almost as good as a kill switch.
There are hundreds of thousands of circumstances where that's completely irrelevant, and they tend to be the situations where you're likely to find an old computer.
I'm running Mac OS 9 on the computer connected to my scanner. It's not connected to anything but my LAN and isn't remotely addressible. If you have an exploit against Mac OS 9, it won't do you any good, you can't get to it. If a store has a non-networked register running Windows 9x or Windows 3.x or MS-DOS 3.0, they don't need to update it, they can keep running it as long as they can get hardware.
There's companies still running NC machines controlled by PDP-11s from the '70s. They can keep running them as long as they can get parts.
"We reject the view," he writes in a letter to the top legal advisor at the Copyright Office, "that copyright owners and their licensees are required to provide consumers with perpetual access to creative works. No other product or service providers are held to such lofty standards. No one expects computers or other electronics devices to work properly in perpetuity, and there is no reason that any particular mode of distributing copyrighted works should be required to do so."
Computers and other products might wear out, but they do not have a "kill switch" that will stop them from working after a specific date, or at the request of the vendor. If you take care of computer hardware, automobiles, other physical objects, they can last a lifetime. The same is true for music, books, and other physical media. DRMed content contains such a "kill switch"... once the server goes down, it's gone.
People used to joke about "having to buy the White Album again", but they didn't actually have to do it, they could keep playing the vinyl copy when CDs came along, and even iTunes didn't make the forty year old LP turn into dust. DRM gives the music industry a new capability, the ability to force EVERYONE to "buy the White Album again" by taking down a single server.
The only functional difference between Bing and Google is that Bing loads a new flash applet every time you move your mouse over a video, and Google waits until you actually click on the video. MUCH friendlier.
Surely the TTY code isn't part of the serial driver subsystem? What the tty subsystem handles is line discipline, and that can be applied to any number of serial interfaces, both physical (serial ports) and virtual (sockets). If you talk to the serial port RAW there shouldn't be a line discipline involved.
(the issue of whether they're in the "kernel" or not is a separate issue, QNX being a microkernel being "in the kernel" there is kind of meaningless)
Memory footprint is a misleading metric. Unused ram is wasted ram.
I'm not talking about cache and other tricks to use free memory for your benefit, I'm talking about how much RAM you need to get decent performance out of the OS. Netbooks tend to be minimally equipped with everything, including RAM, so if you need 1GB to get decent performance out of Vista/7 and 360M to get decent performance out of XP (numbers generated by proctonautics), you're probably going to be happier with XP than Vista/7 on a 1GB netbook, or even a 2GB one.
Let's see... a bunch of hardware benchmarks, which would be expected to result in negligible difference between different versions of Windows. Does Vista REALLY come out significantly worse than XP on these kinds of benchmarks?
How about something relevant to netbooks? What's the memory footprint? Disk footprint?
To me i'd rather get a short preview of the movie before I click on the actual thing rather than actually go to another page, stream the movie and find out its the wrong one.
When you put your mouse over the icons it gives you a preview of the clip along with the sound.
I would rather search for movies by slamming my hand in a door. I hate that kind of laggy UI spam, and turn it off whenever it shows its geeky face. My browser thanks me for it, and rewards me with faster response times wherever I go.
You don't realize how good exchange and outlook are until you don't have them.
I didn't realize how BAD outlook was until I had to use it. I would far rather use Gmail's webmail interface... and I hate the quirky "conversation-oriented" interface I get from Gmail... than Outlook. If I never had to play find-the-option in Outlook, or wait for Outlook to s-l-o-w-l-y search my tiny mailbox (haven't they heard of indexes?), I would be so happy... but they're totally Microsoft-centric at work.
If your IMAP client is freezing, get a better IMAP client. You're not trying to use Outlook for IMAP mail are you, you poor sod?
Unlike Unix, sockets are not a common method of IPC on windows (except in apps written by unix developers). It is a vulnerable design in the first place that two programs on the same machine would communicate in a way that allows network access by default.
Of course they're not using sockets, they're using Lan Manager named pipes. The problem is that once they bound NetBIOS to TCP/IP ports, all the local NetBIOS traffic (and yes, Microsoft services and applications were written to use it internally) was exposed to the Internet.
Isn't that a good thing for the receiving app to re-validate the data?
I'm not talking about *re*validating the data, I'm talking about the requirement for re-quoting and re-parsing atomic command line objects using the same quoting conventions. In UNIX, this only happens when an applications uses the system(3) library function to run an application instead of an exec[lv]-family call. The use of system() has been systematically deprecated for any application that needs to handle untrusted data for decades. In Windows, there's no way to call 'echo "Hello \"World\""' without knowing how echo is going to parse quotes. In UNIX, you call 'execl("/bin/echo", "echo", "Hello\"World\"", NULL);' and you're guaranteed that echo is going to see a single argument, 'Hello "World"', because the API never quotes and reparses it.
It seems to be a different approach to the same problem, with Croquet using distributed synchronization of computation rather than synchronized distribution of updates.
Privilege separation is a useful tool, but minimizing the surface area for the initial attack is critical. Security is like sex, once you're penetrated, you're ****ed.
The biggest problems Windows has are related to the surface area exposed to attack:
1. The lack of the ability to bind most survices to a specific IP address means that even services intended for internal use have to be blocked by a firewall rather than being bound to 127.0.0.1.
2. The lack of ability to pass parameters to a program without passing through a re-parsing step, leading to quoting attacks against helper applications.
3. ActiveX.
4. ActiveX.
5. The use of a common set of helper application bindings for the shell and browser, a vulnerability alas copied by Apple.
6. Did I mention ActiveX?
Windows has privilege separation issues, but not nearly as great as they used to, so I wouldn't put this even in the top 10 security problems.
Common runtimes, like Flash, Silverlight, and Java, are a problem because they create the possibility of a "one size fits all" attack. You shouldn't ignore the danger whether you're running Windows or UNIX.
Underpowered? What the hell is underpowered at dual-quad-core server-hw?
Apple does a good job for their top of the line boxes, but the ones mere humans can afford are positively anemic compared with comparably priced and configured Windows boxes.
Don't get me wrong, I wouldn't give up my Mac to save the 40% or so premium I paid for it over something running Windows. The software... Mac OS X and the rest... is worth every penny of the "Mac Tax"... but pretending the Mac Tax isn't there is wilful blindness.
We have to remember the purpose of IP law - when it ceases to protect creators of intellectual works, it is no longer serving its purpose.
I think that should be "We have to remember the purpose of IP law - when it ceases to encourage the production and publication of intellectual works, it is no longer serving its purpose."
Not that this changes your conclusion significantly... the journal would get a non-exclusive license to the work, rather than copyright, with possibly some embargoed period of exclusivity for the work as a whole. Yes, I know that sounds like it's recreating part of 18th century copyright law as a contract between the author and the publisher, but that may be what's necessary.
Isn't the whole problem of any MMO with a storyline is that you *can't* have an effect on the game world? No matter how often you win, no matter how often you stop the necromancer from drowning puppies, the next set of n00bs to come through has to do the same things and you have to interact with them. If they wanted to do something different in The Matrix, maybe they should have REALLY done something different, put a Thirteenth Floor spin on it and introduced new layers of meta-worlds. Level up far enough, and you wake up again, in the new real "real world"... "Oh, Zion? That was a simulation..."
How can we get anywhere if they are illiterate with brush in hand, or if they can't tell gyronny from verre or gules from sable... these are important skills for the modern scribe or herald!
It's a myth because there's no basis for assuming that Jupiter will throw more junk into wider orbits than into tighter ones more likely to hit the earth. A large planet in the outer solar system is if anything more likely to peturb objects into the inner solar system than out of it. Remember the "Nemesis" theory?
Slashdot Mirror
If a publisher of a non-free computer operating system has announced that it declines to issue more security patches for that version, and you have discovered a remote vulnerability in that operating system version, it's almost as good as a kill switch.
There are hundreds of thousands of circumstances where that's completely irrelevant, and they tend to be the situations where you're likely to find an old computer.
I'm running Mac OS 9 on the computer connected to my scanner. It's not connected to anything but my LAN and isn't remotely addressible. If you have an exploit against Mac OS 9, it won't do you any good, you can't get to it. If a store has a non-networked register running Windows 9x or Windows 3.x or MS-DOS 3.0, they don't need to update it, they can keep running it as long as they can get hardware.
There's companies still running NC machines controlled by PDP-11s from the '70s. They can keep running them as long as they can get parts.
This statement is completely wrong.
"We reject the view," he writes in a letter to the top legal advisor at the Copyright Office, "that copyright owners and their licensees are required to provide consumers with perpetual access to creative works. No other product or service providers are held to such lofty standards. No one expects computers or other electronics devices to work properly in perpetuity, and there is no reason that any particular mode of distributing copyrighted works should be required to do so."
Computers and other products might wear out, but they do not have a "kill switch" that will stop them from working after a specific date, or at the request of the vendor. If you take care of computer hardware, automobiles, other physical objects, they can last a lifetime. The same is true for music, books, and other physical media. DRMed content contains such a "kill switch"... once the server goes down, it's gone.
People used to joke about "having to buy the White Album again", but they didn't actually have to do it, they could keep playing the vinyl copy when CDs came along, and even iTunes didn't make the forty year old LP turn into dust. DRM gives the music industry a new capability, the ability to force EVERYONE to "buy the White Album again" by taking down a single server.
The only functional difference between Bing and Google is that Bing loads a new flash applet every time you move your mouse over a video, and Google waits until you actually click on the video. MUCH friendlier.
Surely the TTY code isn't part of the serial driver subsystem? What the tty subsystem handles is line discipline, and that can be applied to any number of serial interfaces, both physical (serial ports) and virtual (sockets). If you talk to the serial port RAW there shouldn't be a line discipline involved.
(the issue of whether they're in the "kernel" or not is a separate issue, QNX being a microkernel being "in the kernel" there is kind of meaningless)
Memory footprint is a misleading metric. Unused ram is wasted ram.
I'm not talking about cache and other tricks to use free memory for your benefit, I'm talking about how much RAM you need to get decent performance out of the OS. Netbooks tend to be minimally equipped with everything, including RAM, so if you need 1GB to get decent performance out of Vista/7 and 360M to get decent performance out of XP (numbers generated by proctonautics), you're probably going to be happier with XP than Vista/7 on a 1GB netbook, or even a 2GB one.
I speak from experience.
Let's see... a bunch of hardware benchmarks, which would be expected to result in negligible difference between different versions of Windows. Does Vista REALLY come out significantly worse than XP on these kinds of benchmarks?
How about something relevant to netbooks? What's the memory footprint? Disk footprint?
To me i'd rather get a short preview of the movie before I click on the actual thing rather than actually go to another page, stream the movie and find out its the wrong one.
Have you ever actually used Google Video Search?
When you put your mouse over the icons it gives you a preview of the clip along with the sound.
I would rather search for movies by slamming my hand in a door. I hate that kind of laggy UI spam, and turn it off whenever it shows its geeky face. My browser thanks me for it, and rewards me with faster response times wherever I go.
You don't realize how good exchange and outlook are until you don't have them.
I didn't realize how BAD outlook was until I had to use it. I would far rather use Gmail's webmail interface... and I hate the quirky "conversation-oriented" interface I get from Gmail... than Outlook. If I never had to play find-the-option in Outlook, or wait for Outlook to s-l-o-w-l-y search my tiny mailbox (haven't they heard of indexes?), I would be so happy... but they're totally Microsoft-centric at work.
If your IMAP client is freezing, get a better IMAP client. You're not trying to use Outlook for IMAP mail are you, you poor sod?
Isn't making them use Bing bad enough already?
Unlike Unix, sockets are not a common method of IPC on windows (except in apps written by unix developers). It is a vulnerable design in the first place that two programs on the same machine would communicate in a way that allows network access by default.
Of course they're not using sockets, they're using Lan Manager named pipes. The problem is that once they bound NetBIOS to TCP/IP ports, all the local NetBIOS traffic (and yes, Microsoft services and applications were written to use it internally) was exposed to the Internet.
Isn't that a good thing for the receiving app to re-validate the data?
I'm not talking about *re*validating the data, I'm talking about the requirement for re-quoting and re-parsing atomic command line objects using the same quoting conventions. In UNIX, this only happens when an applications uses the system(3) library function to run an application instead of an exec[lv]-family call. The use of system() has been systematically deprecated for any application that needs to handle untrusted data for decades. In Windows, there's no way to call 'echo "Hello \"World\""' without knowing how echo is going to parse quotes. In UNIX, you call 'execl("/bin/echo", "echo", "Hello\"World\"", NULL);' and you're guaranteed that echo is going to see a single argument, 'Hello "World"', because the API never quotes and reparses it.
It seems to be a different approach to the same problem, with Croquet using distributed synchronization of computation rather than synchronized distribution of updates.
I must admit to a healthy dose of schadenfreude when I read about a cheat program getting cracked. :)
Privilege separation is a useful tool, but minimizing the surface area for the initial attack is critical. Security is like sex, once you're penetrated, you're ****ed.
The biggest problems Windows has are related to the surface area exposed to attack:
1. The lack of the ability to bind most survices to a specific IP address means that even services intended for internal use have to be blocked by a firewall rather than being bound to 127.0.0.1.
2. The lack of ability to pass parameters to a program without passing through a re-parsing step, leading to quoting attacks against helper applications.
3. ActiveX.
4. ActiveX.
5. The use of a common set of helper application bindings for the shell and browser, a vulnerability alas copied by Apple.
6. Did I mention ActiveX?
Windows has privilege separation issues, but not nearly as great as they used to, so I wouldn't put this even in the top 10 security problems.
Common runtimes, like Flash, Silverlight, and Java, are a problem because they create the possibility of a "one size fits all" attack. You shouldn't ignore the danger whether you're running Windows or UNIX.
Underpowered? What the hell is underpowered at dual-quad-core server-hw?
Apple does a good job for their top of the line boxes, but the ones mere humans can afford are positively anemic compared with comparably priced and configured Windows boxes.
Don't get me wrong, I wouldn't give up my Mac to save the 40% or so premium I paid for it over something running Windows. The software... Mac OS X and the rest... is worth every penny of the "Mac Tax"... but pretending the Mac Tax isn't there is wilful blindness.
And Tracers was a take-of of Snafu, but our publisher dollied it up with a cyberspace backdrop to make it small like Tron. :(
Light Cycles was just a take-off of the old Intellivision game "SNAFU".
We have to remember the purpose of IP law - when it ceases to protect creators of intellectual works, it is no longer serving its purpose.
I think that should be "We have to remember the purpose of IP law - when it ceases to encourage the production and publication of intellectual works, it is no longer serving its purpose."
Not that this changes your conclusion significantly... the journal would get a non-exclusive license to the work, rather than copyright, with possibly some embargoed period of exclusivity for the work as a whole. Yes, I know that sounds like it's recreating part of 18th century copyright law as a contract between the author and the publisher, but that may be what's necessary.
a package including an LED, lense, and a battery pack?
They've been looking into a scheme using a retroreflector so you can illuminate the code with a camera flash.
That would be "Jupiter as a Sniper Rather Than a Shield"?
I also see "Jupiter: shield or sniper?".
Regarding changing the landscape, it seems that some MMOs definitely do this -- EVE Online comes to mind.
Yah, but isn't EVE pretty much completely PvP (or company vs company), with no formal storyline at all?
Frying pan, meet fire.
I've been reading DRM-free eBooks on my PDA since 2000. There's no reason to spend more on a restricted dedicated book reader instead.
Isn't the whole problem of any MMO with a storyline is that you *can't* have an effect on the game world? No matter how often you win, no matter how often you stop the necromancer from drowning puppies, the next set of n00bs to come through has to do the same things and you have to interact with them. If they wanted to do something different in The Matrix, maybe they should have REALLY done something different, put a Thirteenth Floor spin on it and introduced new layers of meta-worlds. Level up far enough, and you wake up again, in the new real "real world"... "Oh, Zion? That was a simulation..."
How can we get anywhere if they are illiterate with brush in hand, or if they can't tell gyronny from verre or gules from sable... these are important skills for the modern scribe or herald!
It's a myth because there's no basis for assuming that Jupiter will throw more junk into wider orbits than into tighter ones more likely to hit the earth. A large planet in the outer solar system is if anything more likely to peturb objects into the inner solar system than out of it. Remember the "Nemesis" theory?