Ultimately, it'll be pay-per-thought. We'll have cybernetic devices inplanted into the bas of our skulls to meter incoming content. Video/audio can then be sent encrypted all the way to our brains where final and untappable decryption takes place. Even think "Exit light... Enter night! Taaaake my hand! Off to never never land!" and ka-CHING, your credit account is charged a small fee
I think a true test of this model will happen at kuro5hin where Rusty (the kuro5hin (k5) dude) is attempting a model where a user pays $5 a month and, in return, the user will not have to see ads.
The problem with that is/. and K5 ads are the only ones I intentionally don't block, because the OSDN ads are well targeted for geeks like me. I like seeing them. And now with Mozilla.92's working image blocking, I'll never have to punch the monkey or click on the 33 ever again, because the sites those ad images come from are now blocked permanently.
I never SIGNED a contract (has anyone ever had to sign a DSL contract?)
Yes, actually. When I signed up for my DSL back in early 2000, the ISP (a third-party ISP using Southwestern Bell for the link) faxed me a contract, and had me sign it and fax it back. So as of a few months ago, the old but functional Alcatel 1000 should be completely mine. And I even have a tested working spare that I got a couple of months ago for $50 from a thrift store. (there was a bit of paperwork in the box indicating it was used by a customer of another third-party ISP that uses SWB)
As for piles of modems, some people last year did get sent modems two to five times by SWB, and sometimes SWB forgot they sent one of the extras and never asked for it back...
100 meters might not even be enough to get out of the CO, if it has a large enough distribution frame. (The distribution frame is basically an enormous patch panel with the wires to the outside customers and inside equipment on the back side, and trays full of patch wires on the front. Downtown COs in old, large cities can amass quite a mess of patch wires over the years.)
FWIW, I think it uses some bash-specific features like '\$'. Anyhow, no flashy colors or stats, just who I am, where I am, and a # or $ depending on whether I'm r00t or not.
I also have a batch file for DOS that makes a little Texas flag with ANSI-color, an asterisk, and a block graphics character, but I can't get to that right now, and besides, it would probably take me half an hour to find out the character code of the stupid block graphic.
You can have it both ways by posting a mutating but valid address that will expire in a month or so but is still obviously munged such that a human putting it in an address book can figure out your real address. The main ways of doing this are the Sendmail '+' hack (user+arbitrarystring@example.com), and temporary third-level DNS entries (user@7a235f6e.example.com).
Obviously you tried the x86 version of Solaris. Gee, I only paid $15 for Solaris 7 to find out how it sucked. How, you ask? It sucked mostly because it didn't support any video card manufactured in the two years before. In particular, it didn't support any video card I could buy at Best Buy or CompUSA. I had to go to a swap meet to get a card that it would run at better than 640x480x16. CDE looked positively ugly at that resolution. And it wouldn't do text mode beyond 25x80, while Linux usually has no problem with 60x80. I'll stick with Slackware until I have time to look into a BSD distro.
On the other hand, I still keep it around for when I can pick up good obsolete Sparc hardware cheap.
Second of all, how did he get the updated software, unless he was dialing in to TiVo? If he really wanted to use the box without TiVo service, why was he dialing in with the modem? If he was happy with his boat-anchor mode 1.3 box, why not just yank the modem cable out of the back? Why? Because he wanted the clock set by TiVo, so, he basically wanted to dial-in to their service for free.
RTFL. Or at least RTFL more carefully. He didn't have it dial in because he was too lazy to set the clock himself. He had it dial in because he couldn't set the clock himself. TiVo doesn't let you set the clock on your own. You have to dial in! If it weren't for this one little camel's nose under the edge of the tent, he wouldn't have had it dial in.
At this point, I'm glad I never got around to getting one of these things. This, plus the other thing I've heard they've tried (putting a commercial on the screen during pause) shows that they care more about money than their customers.
Great idea. Now where are you going to get the copper to go from a cell phone tower hut to the various houses in the neighborhood? Unless the tower was built right on top of a major neighborhood copper cable, you're out of luck.
Intermittent connections, store-and-forward bundles, scheduled connection times, that's a lot like FidoNet (and uucp). The asymmetrical bandwidth is the only real difference.
Ah yes, back to the good old days of e-mail, usenet, and ftp-mail. Just imagine trying to browse the web (or even just trying to use Slashdot!) over e-mail.
I can remember back in the early '90s how my only access to a very few select newsgroups was through a gateway to Fidonet, with my BBS (which wasn't really intended for users--they would just hog the connection) dialing long distance at midnight to get the feed.
The ATI Rage in my Pismo (Powerbook G3 Firewire) does this as well. So the result of viewing 800x600 on a 1024x768 display is a bit fuzzy, but it's not that horrid chunky effect you get from LCD monitors that run off of analog SVGA inputs.
I'm pretty sure that these monitors that Apple will be selling do not have a VGA connector, but instead have a direct digital interface. That means that the video chip will know what resolution of LCD is being driven, and can interpolate-scale the picture accordingly.
I think this is a good move by Apple. The margins on CRT monitors are probably not all that great, and there are a lot of competing SVGA monitors available. I have two monitors on the tower Mac back at my apartment, and neither one has a picture of a fruit on it.
I can't believe all the whining about "but I'm a graphic artist and LCD doesn't give reliable colors!" Well, DUHHH, you don't have to be a mind numbed robot and only buy monitors from the same company that made your computer! And if you only use Pantone colors, it doesn't really matter how accurate the monitor is; you should be picking colors out of your swatch book.
I am worried about one thing, though. Since these monitors will all be direct digital, that brings us one step closer to the perfect world as defined by the MPAA and RIAA, where everything will be encrypted right up to the display and speakers.
I just checked my logs, and I've only had three hits on 10008. One from Canada, another from Korea, and the third from Sweden. That was three days ago, in a six hour period. So at least it doesn't look like this thing is going to melt down the internet.
I'm running 0.9 under W2K, and I noticed in Task Manager that it was taking up about 75 megs of process space, bringing my commit charge really close to the 256 megs of RAM in the machine. I hit minimize and it dropped to 2 (!) megs, then un-minimized and it went up to about 7. And right now it's back up to 35 megs. That's a significant difference!
It must leave all sorts of pre-cached data lying around when it's active, and tosses it when minimized. Pretty cool.
Which 'sploit do you choose?
on
Cracking OSX
·
· Score: 2
So tell me, how do you 'sploit two different architectures at once, when a given remote root buffer overflow exploit has only one chance because they crash the vulnerable task if they fail to take it over?
Think about it... if you tried to hit an OS X machine with a Linux BIND 'sploit, it would crash the nameserver for a simple DoS. But because BIND has now crashed, it's no longer vulnerable! No root for you!
You better get your OS fingerprinting right the first time through! Seems with appropriate fingerprinting obfuscation, OS X would be a nice choice for running a BIND nameserver.
Of course inetd daemons like wuftpd and sendmail let you have another chance, but I doubt OS X is running wuftpd, and sendmail is pretty clean these days. And so is apache.
By the way, because the PowerPC is big-endian, there is a class of buffer exploits that it is immune to. Sometimes only one byte of buffer overflow is possible, and on an X86 that lets you change the low byte of the return address to return to a wrong place near the original caller. On the PPC, that's now the high byte of the return address, which is basically useless.
"War driving", huh? I've done that a couple of times myself just to see what I'd find, and while I saw more than few SSIDs (including a couple on the same residential street where the SSID names indicated AirPort users), I think I only ever found one that was truly wide-open.
These guys in the article only mention getting into maybe two or three networks, but they see a lot more. A visible SSID isn't necessarily bad security by itself, and it's a hell of a lot better than a wide-open 802.11b that lets you surf the web from somebody else's parking lot!
Not much better is a building with live DHCP'ed 100-base-T ports in every room. I've found at least one university building like this. I wonder how long until people hiding base stations (not advertising SSIDs, of course) in the ceiling spaces of buildings gets to be a real problem.
Hydrogen was not what downed the Hindenburg. This is a popular misconception. The real problem was that the damn thing had a reflective coating of powdered aluminum, so that the hydrogen wouldn't get too warm from sunlight or something... esentially, it was coated with Thermite! All it took was one spark, and POOF!
So what about our favorite "viral license", the GPL? If this is a loadable kernel module ONLY, they can get around it. But if they link with the kernel, they have to release source code, which I'm sure they do NOT want to do.
On the other hand, if it's just a LKM, then it's in a nice self-contained file where it's easier to disassemble and patch, or even replace with code that does nothing.
Macrovision doesn't mess with timing at all. It puts a spike signal into the vertical retrace. This messes with signal levels in such a way that the cheap AGC (automatic gain control) circuits in most VHS video recorders freak out and mess with the brightness of the picture. There's also a Macrovision II (aka Colorstripe) which messes with the colorburst signal on a couple of lines of the picture, but this can't even be recorded onto master VHS tapes, and is only generated by DVD players. Either way, since this is patented by Macrovision, studios must pay MV whenever they enable either of these features on a DVD disc.
Betamax video recorders were not vulnerable to this signal and would supposedly copy it perfectly. Ditto for Go Video VHS recorders, although I think they were eventually forced to put the crappy AGC circuit into their stuff a couple of years back.
Time Base Correctors fix the problem because the nasty signals are in the sync areas of the screen, which they throw out and replace with clean signals. But a true TBC (which makes sure that entire scanlines come through with the right timing, something that videotape is not accurate at) is overkill for Macrovision.
Slashdot Mirror
I believe this is what you're referring to?
The problem with that is /. and K5 ads are the only ones I intentionally don't block, because the OSDN ads are well targeted for geeks like me. I like seeing them. And now with Mozilla .92's working image blocking, I'll never have to punch the monkey or click on the 33 ever again, because the sites those ad images come from are now blocked permanently.
Yes, actually. When I signed up for my DSL back in early 2000, the ISP (a third-party ISP using Southwestern Bell for the link) faxed me a contract, and had me sign it and fax it back. So as of a few months ago, the old but functional Alcatel 1000 should be completely mine. And I even have a tested working spare that I got a couple of months ago for $50 from a thrift store. (there was a bit of paperwork in the box indicating it was used by a customer of another third-party ISP that uses SWB)
As for piles of modems, some people last year did get sent modems two to five times by SWB, and sometimes SWB forgot they sent one of the extras and never asked for it back...
100 meters might not even be enough to get out of the CO, if it has a large enough distribution frame. (The distribution frame is basically an enormous patch panel with the wires to the outside customers and inside equipment on the back side, and trays full of patch wires on the front. Downtown COs in old, large cities can amass quite a mess of patch wires over the years.)
PS1='\u@\h:\w\$ '
FWIW, I think it uses some bash-specific features like '\$'. Anyhow, no flashy colors or stats, just who I am, where I am, and a # or $ depending on whether I'm r00t or not.
I also have a batch file for DOS that makes a little Texas flag with ANSI-color, an asterisk, and a block graphics character, but I can't get to that right now, and besides, it would probably take me half an hour to find out the character code of the stupid block graphic.
You can have it both ways by posting a mutating but valid address that will expire in a month or so but is still obviously munged such that a human putting it in an address book can figure out your real address. The main ways of doing this are the Sendmail '+' hack (user+arbitrarystring@example.com), and temporary third-level DNS entries (user@7a235f6e.example.com).
On the other hand, I still keep it around for when I can pick up good obsolete Sparc hardware cheap.
Finally, a replacement for "ALL YOUR BASE ARE BELONG TO US"!
Yeah, maybe you'll make Jimmy Carter smile.
RTFL. Or at least RTFL more carefully. He didn't have it dial in because he was too lazy to set the clock himself. He had it dial in because he couldn't set the clock himself. TiVo doesn't let you set the clock on your own. You have to dial in! If it weren't for this one little camel's nose under the edge of the tent, he wouldn't have had it dial in.
At this point, I'm glad I never got around to getting one of these things. This, plus the other thing I've heard they've tried (putting a commercial on the screen during pause) shows that they care more about money than their customers.
I'll wait for a more "open" solution.
Great idea. Now where are you going to get the copper to go from a cell phone tower hut to the various houses in the neighborhood? Unless the tower was built right on top of a major neighborhood copper cable, you're out of luck.
Ah yes, back to the good old days of e-mail, usenet, and ftp-mail. Just imagine trying to browse the web (or even just trying to use Slashdot!) over e-mail.
I can remember back in the early '90s how my only access to a very few select newsgroups was through a gateway to Fidonet, with my BBS (which wasn't really intended for users--they would just hog the connection) dialing long distance at midnight to get the feed.
I'm pretty sure that these monitors that Apple will be selling do not have a VGA connector, but instead have a direct digital interface. That means that the video chip will know what resolution of LCD is being driven, and can interpolate-scale the picture accordingly.
I think this is a good move by Apple. The margins on CRT monitors are probably not all that great, and there are a lot of competing SVGA monitors available. I have two monitors on the tower Mac back at my apartment, and neither one has a picture of a fruit on it.
I can't believe all the whining about "but I'm a graphic artist and LCD doesn't give reliable colors!" Well, DUHHH, you don't have to be a mind numbed robot and only buy monitors from the same company that made your computer! And if you only use Pantone colors, it doesn't really matter how accurate the monitor is; you should be picking colors out of your swatch book.
I am worried about one thing, though. Since these monitors will all be direct digital, that brings us one step closer to the perfect world as defined by the MPAA and RIAA, where everything will be encrypted right up to the display and speakers.
I just checked my logs, and I've only had three hits on 10008. One from Canada, another from Korea, and the third from Sweden. That was three days ago, in a six hour period. So at least it doesn't look like this thing is going to melt down the internet.
What I want to know is why "Estados Unidos" (referring to the USA) is abbreviated "EEUU".
ARGH! I read that just as I was putting the cola can up to my lips. It could be worse, I could have had my mouth full of cola when I read it!
It must leave all sorts of pre-cached data lying around when it's active, and tosses it when minimized. Pretty cool.
Think about it... if you tried to hit an OS X machine with a Linux BIND 'sploit, it would crash the nameserver for a simple DoS. But because BIND has now crashed, it's no longer vulnerable! No root for you!
You better get your OS fingerprinting right the first time through! Seems with appropriate fingerprinting obfuscation, OS X would be a nice choice for running a BIND nameserver.
Of course inetd daemons like wuftpd and sendmail let you have another chance, but I doubt OS X is running wuftpd, and sendmail is pretty clean these days. And so is apache.
By the way, because the PowerPC is big-endian, there is a class of buffer exploits that it is immune to. Sometimes only one byte of buffer overflow is possible, and on an X86 that lets you change the low byte of the return address to return to a wrong place near the original caller. On the PPC, that's now the high byte of the return address, which is basically useless.
These guys in the article only mention getting into maybe two or three networks, but they see a lot more. A visible SSID isn't necessarily bad security by itself, and it's a hell of a lot better than a wide-open 802.11b that lets you surf the web from somebody else's parking lot!
Not much better is a building with live DHCP'ed 100-base-T ports in every room. I've found at least one university building like this. I wonder how long until people hiding base stations (not advertising SSIDs, of course) in the ceiling spaces of buildings gets to be a real problem.
I second that. The President's Analyst should be on every list of Must-See Geek Movies, right up there with Real Genius. James Coburn rules.
What's next... Dungeons & Dragons?
Hydrogen was not what downed the Hindenburg. This is a popular misconception. The real problem was that the damn thing had a reflective coating of powdered aluminum, so that the hydrogen wouldn't get too warm from sunlight or something... esentially, it was coated with Thermite! All it took was one spark, and POOF!
On the other hand, if it's just a LKM, then it's in a nice self-contained file where it's easier to disassemble and patch, or even replace with code that does nothing.
Betamax video recorders were not vulnerable to this signal and would supposedly copy it perfectly. Ditto for Go Video VHS recorders, although I think they were eventually forced to put the crappy AGC circuit into their stuff a couple of years back.
Time Base Correctors fix the problem because the nasty signals are in the sync areas of the screen, which they throw out and replace with clean signals. But a true TBC (which makes sure that entire scanlines come through with the right timing, something that videotape is not accurate at) is overkill for Macrovision.
1904 was chosen on the Mac because 1900 was not a leap year, and it made the calculation easier.