Here in Germany we only pay for making the calls. The receiving end doesn't get charged at all.
So the caller gets overcharged for cellular calls, whichever end he's on. That's fine, it's overcharging a different person than in the US, but whatever...
Either way if I'm on a cellphone, and you're on a cellphone, and I call you... either I get charged for both sets of airtime (like in.de) or we each get charged for our own airtime (like in.us). But only one person pays for the airtime.
What AT&T seems to want to do is charge Google extra for being Google, even though Google is already "paying for their airtime".
What's the point of "For example, what happens if someone wants to do a policy where there's CD copy protection, but after the first month [consumers] can download a patch that gets rid of it."?
I don't understand what the company would imagine they would get out of that. The logic would be "If you can disable the copy protection, then as soon as the first consumer can do so, what's to stop that person from sharing the patch, or the patched copy." The only way that would work would be if they're not actually "removing the DRM", they're just changing what it's locked to. Why would he see that as "legitimate"?
The "Tier 1" carriers charge the ISP on the client end of a connection, and they also charge the ISP on the server end on the connection.
I know. I already pointd that out myself. In the very message that you just responded to.
If you feel like traffic should only be charged in one direction
I didn't, however, say that.
I'm talking about the frequently raised argument that they should be able to charge Google (or whoever) again, via some kind of "guaranteed service" (without which they'll of course 'unguarantee' the service they're currently getting) on top of everything else.
If you think it's easy and cheap to be a tier 1 carrier,
Where did I say that? I didn't.
or that it makes more sense to only charge half your customers, [...]
They're not "only charging half their customers", though. They're getting paid for all their downstream connections. What they theoretically miss when they cross-charge for peering is paid for by their actual customers on their downstreams, or by reduced charges at other peering points.
Google pays for their traffic into and out of their POP. My ISP pays for traffic into and out of its POP. If AT&T's not getting paid directly by Google, they're getting paid (directly or indirectly) by whoever they peer with that handles Google's POP. There's no point where all the traffic isn't already being paid for by someone.
AT&T wants to charge Google directly, as well as charging Google's ISP. That's double-dipping, because they're already getting paid for that traffic.
Ya know, when I ask Linux "How much can this 1TB drive hold?" and it reports back "900 GB".
And when you copy files to it, you find that it only holds 760GB, because your files are small enough that the file system overhead eats more than 10% of the available space.
Gibibyte is dead because the difference between 2^30 octets and 10^9 octets is small. The computer industry uses Gigabyte for 2^30 octets because it works in powers of two, so the storage PART of the industry should do the same thing.
OK, I don't know if Arthur C Clarke or Robert Heinlein or Isaac Asimov or Frank Herbert invented the iPod too, but I would be absolutely shocked if there wasn't something in a SF story from the '50s or '60s that looked like a handheld digital music player. Hell, you could probably find one in the original Star Trek.
"Computer... Shuffle... The White Album, Dark Side of the Moon, and Hotel California."
To my mind, calling 1,024 bytes a 'kilobyte' was just about acceptable, since the difference wasn't so great and 'kilo' was a convenient shorthand. But calling 1,073,741,824 bytes a 'gigabyte' is really pushing it,
Why? If you have a 100 GB drive, and you have 90GB of files to put into it, you can't tell whether those files will fit unless you know the file system you're putting them into, how many files you have, how the file sizes are distributed, because the file system overhead on those files can be more than the total sizes of the files put together if they're small enough... and it's not at all unusual for it to be 10-20%. The 7% difference between 10^9 and 2^30 is lost in the noise.
Why do I give a crap about what kind of arithmetic the COMPUTER uses?
Because the only time it matters whether it's 2^30 bytes or 10^9 bytes is when you're doing something that really truly does matter to the computer.
I can say "oh, somefile.dat is about 3.4 GB".
Which is still useless information, because your file system overhead on that file may be anything from 1% to over 10% depending on the file system. If it's in a bunch of small files, it's even worse... the file system overhead on a directory tree can be half as much again as the files inside it. In general, if you have 90 GB of files, and a 100 GB drive, you can't tell whether the files are going to fit or not... whether the directory listing shows "90000000000 bytes", "90000000 kB", or "90000 MB" or even "90 GB"... and whether those are powers-of-two or powers-of-10 prefixes.
Or alternatively we need RAM manufacturers to stop defining 'gigabyte' as '1,073,741,824 bytes'.
It's not RAM manufacturers, it's the whole computer industry... and for a good reason, that being that computers haven't used decimal arithmetic since COBOL was new and sexy.
Nobody uses 'GiB'. It was a fad, and it's a dead fad. And in any case it should be 'Gio'... the 8-bit-byte is actually LESS of a standard than the 2^30 octet Go.
But they don't charge the guy on the other end of the call airtime as well. What these guys are talking about amounts to charging BOTH ends of EVERY cellular call, no matter who made it, ALL the airtime charges.
Is that why I pay for my incoming SMS and calls?;)
It's why I don't pay for calling you if you have a cellphone and I don't. And why I don't pay more for RECEIVING a call from a cellphone. Or, when YOU'RE calling a cellphone, it doesn't cost you twice as much as if you're calling a landline phone.
They want to do the equivalent of charging people calling you for your airtime, and charging you for your airtime as well. If you do the equivalent of making a call from a cellphone to another cellphone, they want to charge you twice.
At least the cellphone carriers only overcharge *one* end of the conversation for airtime. These beggers are arguing that they should get to charge BOTH ends the full price of the traffic.
This sounds like the funkiest ad since the "Infiniti" intro that David Horsey skewered so effectively in "Boomer's Song"... Stanley Doobie bought an "Absurditi" and ended up with a Bonsai plant that represented "the idea of the car".
Vista isn't an operating system... it's the *idea* of an operating system.
Regardless of "fairness", why "don't they like" bugmenot badly enough to ban it?
What are they afraid of? Bugmenot explicitly doesn't support Facebook, so it can't be that they're concerned about people using bugmenot, so what's the point?
Google has certificates of trust, and has "Validated EXE files". Part of YOUR security settings, is to allow a "CLICK" as confirmation to a download.
You can tell your browser not to allow downloading from "Certified" and "Validated"... but than you will not get any updates, or even be prompted for any of those programs that "Expect" that since they paid for the "Certification" and "Validation"... and you select NOT to allow... they assume you don't want it, and never prompt.
First of all, Read The Fine Followups. Because it's not whatever you're talking about up above,it's something called the Google Update Service. And it can, in fact, download and install a program without clicking anything.
Second, no browser should ever do what you described above, ever, and Firefox (at least) doesn't do that, ever. What you're describing sounds like something Microsoft would come up with, yet another part of the toxic swamp that is Active X.
Not only that, but Chrome and IE 8 both support ActiveX, which has been the biggest security problem on Windows over the past decade.
And you don't even need to install Chrome to take advantage of Google's security enhancements, if you've installed any other Google Apps the "Google Update" browser API allows eb pages to not just download but *install* software without any user interaction.
Multiple processes need not be expensive.
on
Chrome Vs. IE 8
·
· Score: 2, Interesting
If multiple processes are particularly expensive in Chrome and IE8, that's a problem with Chrome and IE8... or a problem with Windows. At the very least, multiple processes doesn't mean duplicating *everything*... there's no reason to have all the possible plugins and all the web controls and access methods loaded and initialized in all tabs... in fact NOT having that overhead in the context of every tab should be a significant advantage of the design.
Maybe that's the problem. As I see it if someone takes FreBSD but modifies it it's still FreeBSD. Unless the mods are extensive.
The modifications are more than "extensive". It's a completely different OS design, from the ground up. The scheduler is different, the virtual memory is different, the interprocess communication is different, the file systems are different, the driver model is different, the boot process is different... it's like two completely different vehicles built from some of the same components. It's based on FreeBSD (and Mach, and NeXTstep, and Mac OS), but it's not FreeBSD. It's a completely different approach to OS design from the ground up.
There really are more similarities between FreeBSD, Linux, and SCO UNIX than there are between Darwin and FreeBSD. FreeBSD is a monolithic kernel, like Linux, like SCO, like Version 7. Darwin is a services- or subsystems-based kernel, like Tru64 UNIX, or even Windows NT.
There are more differences that I haven't gone into... for example FreeBSD is built from a unitary source tree, like traditional UNIX. Darwin is built like Linux, out of a collection of packages. FreeBSD's system software is organized in a traditional UNIX filesystem hierarchy... with/usr,/var, and so on, Darwin only uses that for the BSD subsystem and loads all the rest of the subsystems from/System and stores configurations in/Library.
PS: The big issue is Google Updater security
on
Google Chrome, Day 2
·
· Score: 1
Yes, several people have noted that it's Google Updater.
Now that we've established that, why does it have an API that allows a web page to request the unattended installation of software on your computer. That's an insane security risk, even if their security model is sound (after all, they are providing even less protection than ActiveX), and we don't know if the security model is good because neither the API nor the security model is documented.
Oh, speaking of ActiveX, apparently Google Chrome includes a shim to allow ActiveX components to execute. That's two really bad ideas for the price of one.
Since Chrome isn't a service, by the letter of Google's EULA for chrome, you're not using their service so they don't won your work, BUT they own every website you view (display) using it.
I'm sorry, but what? You're complaining because some Google software that you're aware of, warned you of what it (c|w)ould do, you agreed to it, and then it did it?
No.
First, it doesn't matter what I did or did not agree to. What they actually implemented appears to be an even bigger backdoor than ActiveX as it currently exists. It is a dangerous privilege escalation for ANY web page, no matter where it's coming from, to initiate the installation of any software. It is unacceptable for that to happen
Second, I do not mean "I recall, even vaguely, doing so", I mean "I am sure that Google has their asses covered". It took quite a bit of searching to find the program that I'd installed that included Google Update.
Third, whatever put in their agreement would not have said "this application will allow Google to install new software on your computer without any action by you". The person who wrote the document may not even have been aware that the API they're using has this capability (which it does, the code I quoted would have worked just as well from OnLoad as from a click).
Google Update Service... backdoor?
on
Google Chrome, Day 2
·
· Score: 2, Informative
OK, it's using a firefox extension called "Google Update", which is installed by a number of Google applications, and (as demonstrated) it's possible for Google to use it to automatically install software on your computer. Disabling it in Firefox keeps it from running and should hopefully prevent some other Google App from installing it again.
This seems to be the same kind of "trust me" backdoor as the Firefox XPI installer and ActiveX, but unlike ActiveX it's cross-browser (and probably cross-platform), and unlike XPI it DOESN'T require you to explicitly whitelist each site and approve each install. Calling it an updater sort of implies that it is a relatively secure service, like other update services, that only pulls down and updates software that you have explicitly installed. But if it has a mechanism for a web site to explicitly request that a new component be downloaded and installed it's anything but secure.
Googling for it on "site:google.com" has been less than useful. I've got several hits from people in Google's user groups asking what it is and how to remove it, but there doesn't seem to be any documentation on Google's website for its API and security model.
Slashdot Mirror
Here in Germany we only pay for making the calls. The receiving end doesn't get charged at all.
So the caller gets overcharged for cellular calls, whichever end he's on. That's fine, it's overcharging a different person than in the US, but whatever...
Either way if I'm on a cellphone, and you're on a cellphone, and I call you... either I get charged for both sets of airtime (like in .de) or we each get charged for our own airtime (like in .us). But only one person pays for the airtime.
What AT&T seems to want to do is charge Google extra for being Google, even though Google is already "paying for their airtime".
What's the point of "For example, what happens if someone wants to do a policy where there's CD copy protection, but after the first month [consumers] can download a patch that gets rid of it."?
I don't understand what the company would imagine they would get out of that. The logic would be "If you can disable the copy protection, then as soon as the first consumer can do so, what's to stop that person from sharing the patch, or the patched copy." The only way that would work would be if they're not actually "removing the DRM", they're just changing what it's locked to. Why would he see that as "legitimate"?
The "Tier 1" carriers charge the ISP on the client end of a connection, and they also charge the ISP on the server end on the connection.
I know. I already pointd that out myself. In the very message that you just responded to.
If you feel like traffic should only be charged in one direction
I didn't, however, say that.
I'm talking about the frequently raised argument that they should be able to charge Google (or whoever) again, via some kind of "guaranteed service" (without which they'll of course 'unguarantee' the service they're currently getting) on top of everything else.
And it's not just AT&T that have raised this.
If you think it's easy and cheap to be a tier 1 carrier,
Where did I say that? I didn't.
or that it makes more sense to only charge half your customers, [...]
They're not "only charging half their customers", though. They're getting paid for all their downstream connections. What they theoretically miss when they cross-charge for peering is paid for by their actual customers on their downstreams, or by reduced charges at other peering points.
Google pays for their traffic into and out of their POP. My ISP pays for traffic into and out of its POP. If AT&T's not getting paid directly by Google, they're getting paid (directly or indirectly) by whoever they peer with that handles Google's POP. There's no point where all the traffic isn't already being paid for by someone.
AT&T wants to charge Google directly, as well as charging Google's ISP. That's double-dipping, because they're already getting paid for that traffic.
Ya know, when I ask Linux "How much can this 1TB drive hold?" and it reports back "900 GB".
And when you copy files to it, you find that it only holds 760GB, because your files are small enough that the file system overhead eats more than 10% of the available space.
Gibibyte is dead because the difference between 2^30 octets and 10^9 octets is small. The computer industry uses Gigabyte for 2^30 octets because it works in powers of two, so the storage PART of the industry should do the same thing.
OK, I don't know if Arthur C Clarke or Robert Heinlein or Isaac Asimov or Frank Herbert invented the iPod too, but I would be absolutely shocked if there wasn't something in a SF story from the '50s or '60s that looked like a handheld digital music player. Hell, you could probably find one in the original Star Trek.
"Computer... Shuffle... The White Album, Dark Side of the Moon, and Hotel California."
a tebibyte is 10% more than a terabyte, and that's almost 100GB - quite a substantial difference.
100GB out of 1TB is lost in the noise.
To my mind, calling 1,024 bytes a 'kilobyte' was just about acceptable, since the difference wasn't so great and 'kilo' was a convenient shorthand. But calling 1,073,741,824 bytes a 'gigabyte' is really pushing it,
Why? If you have a 100 GB drive, and you have 90GB of files to put into it, you can't tell whether those files will fit unless you know the file system you're putting them into, how many files you have, how the file sizes are distributed, because the file system overhead on those files can be more than the total sizes of the files put together if they're small enough... and it's not at all unusual for it to be 10-20%. The 7% difference between 10^9 and 2^30 is lost in the noise.
Why do I give a crap about what kind of arithmetic the COMPUTER uses?
Because the only time it matters whether it's 2^30 bytes or 10^9 bytes is when you're doing something that really truly does matter to the computer.
I can say "oh, somefile.dat is about 3.4 GB".
Which is still useless information, because your file system overhead on that file may be anything from 1% to over 10% depending on the file system. If it's in a bunch of small files, it's even worse... the file system overhead on a directory tree can be half as much again as the files inside it. In general, if you have 90 GB of files, and a 100 GB drive, you can't tell whether the files are going to fit or not... whether the directory listing shows "90000000000 bytes", "90000000 kB", or "90000 MB" or even "90 GB"... and whether those are powers-of-two or powers-of-10 prefixes.
Or alternatively we need RAM manufacturers to stop defining 'gigabyte' as '1,073,741,824 bytes'.
It's not RAM manufacturers, it's the whole computer industry... and for a good reason, that being that computers haven't used decimal arithmetic since COBOL was new and sexy.
Nobody uses 'GiB'. It was a fad, and it's a dead fad. And in any case it should be 'Gio'... the 8-bit-byte is actually LESS of a standard than the 2^30 octet Go.
But they don't charge the guy on the other end of the call airtime as well. What these guys are talking about amounts to charging BOTH ends of EVERY cellular call, no matter who made it, ALL the airtime charges.
Is that why I pay for my incoming SMS and calls? ;)
It's why I don't pay for calling you if you have a cellphone and I don't. And why I don't pay more for RECEIVING a call from a cellphone. Or, when YOU'RE calling a cellphone, it doesn't cost you twice as much as if you're calling a landline phone.
They want to do the equivalent of charging people calling you for your airtime, and charging you for your airtime as well. If you do the equivalent of making a call from a cellphone to another cellphone, they want to charge you twice.
At least the cellphone carriers only overcharge *one* end of the conversation for airtime. These beggers are arguing that they should get to charge BOTH ends the full price of the traffic.
This sounds like the funkiest ad since the "Infiniti" intro that David Horsey skewered so effectively in "Boomer's Song"... Stanley Doobie bought an "Absurditi" and ended up with a Bonsai plant that represented "the idea of the car".
Vista isn't an operating system... it's the *idea* of an operating system.
You can have ZFS. You just need to use an OS that's got a license that doesn't have insane compatibility requirements.
Regardless of "fairness", why "don't they like" bugmenot badly enough to ban it?
What are they afraid of? Bugmenot explicitly doesn't support Facebook, so it can't be that they're concerned about people using bugmenot, so what's the point?
Maybe they could call it Objectivism.
Or Atlas.
*shrug*
Google has certificates of trust, and has "Validated EXE files". Part of YOUR security settings, is to allow a "CLICK" as confirmation to a download.
You can tell your browser not to allow downloading from "Certified" and "Validated"... but than you will not get any updates, or even be prompted for any of those programs that "Expect" that since they paid for the "Certification" and "Validation"... and you select NOT to allow... they assume you don't want it, and never prompt.
First of all, Read The Fine Followups. Because it's not whatever you're talking about up above,it's something called the Google Update Service. And it can, in fact, download and install a program without clicking anything.
Second, no browser should ever do what you described above, ever, and Firefox (at least) doesn't do that, ever. What you're describing sounds like something Microsoft would come up with, yet another part of the toxic swamp that is Active X.
Not only that, but Chrome and IE 8 both support ActiveX, which has been the biggest security problem on Windows over the past decade.
And you don't even need to install Chrome to take advantage of Google's security enhancements, if you've installed any other Google Apps the "Google Update" browser API allows eb pages to not just download but *install* software without any user interaction.
If multiple processes are particularly expensive in Chrome and IE8, that's a problem with Chrome and IE8... or a problem with Windows. At the very least, multiple processes doesn't mean duplicating *everything*... there's no reason to have all the possible plugins and all the web controls and access methods loaded and initialized in all tabs... in fact NOT having that overhead in the context of every tab should be a significant advantage of the design.
Maybe that's the problem. As I see it if someone takes FreBSD but modifies it it's still FreeBSD. Unless the mods are extensive.
The modifications are more than "extensive". It's a completely different OS design, from the ground up. The scheduler is different, the virtual memory is different, the interprocess communication is different, the file systems are different, the driver model is different, the boot process is different... it's like two completely different vehicles built from some of the same components. It's based on FreeBSD (and Mach, and NeXTstep, and Mac OS), but it's not FreeBSD. It's a completely different approach to OS design from the ground up.
There really are more similarities between FreeBSD, Linux, and SCO UNIX than there are between Darwin and FreeBSD. FreeBSD is a monolithic kernel, like Linux, like SCO, like Version 7. Darwin is a services- or subsystems-based kernel, like Tru64 UNIX, or even Windows NT.
There are more differences that I haven't gone into... for example FreeBSD is built from a unitary source tree, like traditional UNIX. Darwin is built like Linux, out of a collection of packages. FreeBSD's system software is organized in a traditional UNIX filesystem hierarchy... with /usr, /var, and so on, Darwin only uses that for the BSD subsystem and loads all the rest of the subsystems from /System and stores configurations in /Library.
Yes, several people have noted that it's Google Updater.
Now that we've established that, why does it have an API that allows a web page to request the unattended installation of software on your computer. That's an insane security risk, even if their security model is sound (after all, they are providing even less protection than ActiveX), and we don't know if the security model is good because neither the API nor the security model is documented.
Oh, speaking of ActiveX, apparently Google Chrome includes a shim to allow ActiveX components to execute. That's two really bad ideas for the price of one.
Since Chrome isn't a service, by the letter of Google's EULA for chrome, you're not using their service so they don't won your work, BUT they own every website you view (display) using it.
I'm sorry, but what? You're complaining because some Google software that you're aware of, warned you of what it (c|w)ould do, you agreed to it, and then it did it?
No.
First, it doesn't matter what I did or did not agree to. What they actually implemented appears to be an even bigger backdoor than ActiveX as it currently exists. It is a dangerous privilege escalation for ANY web page, no matter where it's coming from, to initiate the installation of any software. It is unacceptable for that to happen
Second, I do not mean "I recall, even vaguely, doing so", I mean "I am sure that Google has their asses covered". It took quite a bit of searching to find the program that I'd installed that included Google Update.
Third, whatever put in their agreement would not have said "this application will allow Google to install new software on your computer without any action by you". The person who wrote the document may not even have been aware that the API they're using has this capability (which it does, the code I quoted would have worked just as well from OnLoad as from a click).
OK, it's using a firefox extension called "Google Update", which is installed by a number of Google applications, and (as demonstrated) it's possible for Google to use it to automatically install software on your computer. Disabling it in Firefox keeps it from running and should hopefully prevent some other Google App from installing it again.
This seems to be the same kind of "trust me" backdoor as the Firefox XPI installer and ActiveX, but unlike ActiveX it's cross-browser (and probably cross-platform), and unlike XPI it DOESN'T require you to explicitly whitelist each site and approve each install. Calling it an updater sort of implies that it is a relatively secure service, like other update services, that only pulls down and updates software that you have explicitly installed. But if it has a mechanism for a web site to explicitly request that a new component be downloaded and installed it's anything but secure.
Googling for it on "site:google.com" has been less than useful. I've got several hits from people in Google's user groups asking what it is and how to remove it, but there doesn't seem to be any documentation on Google's website for its API and security model.