I believe that disclosing how your site is performing in that way is against the Google AdSense Terms and Conditions (see clause 7). You can give the total amount of payment, but you've also given the number of clicks you get, which the T&Cs disallow.
I find it interesting that nobody seems to have focused on one thing in particular yet...
The secret message explaining how to redeem the free sandwich offer is planted in just a few of the spot's frames and revealed when replayed in slow motion.
"To the naked eye, it's a typical KFC television commercial," he said.
So tell me, how is this different from subliminal messages? Of course, in this case you know that you're looking for a secret code. But the very practice of hiding messages in a frame or two is the exact same one as used by people looking to put subliminal messages in. If this isn't treated as subliminal messaging, then it could blur the line.
Disclaimer: I wasn't able to watch the video for some reason. I only heard audio. So I can't tell how obvious it actually is.
I'm sure Rogers would love to see your foolproof solution. I mean, you obviously have one.
Seriously, I agree with your reasoning for most things. And yes, this definitely is security through obscurity, which is a bad thing. However, how else would you do an automatic fraud detection system on phone calls? Of course, you could always give the (probably correct) response that automatic fraud detection is a bad thing as it's never going to be perfect, and there will always be those who slip through, as in this case.
Much as I hate to admit it, though, I don't know how else it could be implemented. Fraud is a serious thing that needs to be watched for, and there would be an outcry if Rogers removed it.
It's p2p, yes. But p2p needs some servers to search with. Assuming they can get their new software onto the SuperNodes (if they use different terminology now, please correct me), the server would be able to block searches no problem. It might be hit-and-miss sometimes, though.
Exploiting a flaw in this decoding is as simple as exploiting any other software bug.
Perhaps - except that according to the Wikipedia entry on the Red Book standard, it's stored in PCM format, which to my knowledge can be passed straight to the soundcard with no further processing. The only decoding in the procedure is done by the CD drive and the soundcard, and while exploiting a flaw in either of those might not be difficult, it would also be highly dependent on the chipset used, and in the end would probably both not be able to do much in the grand scheme of things, *and* would probably cause other CD drives/soundcards to barf.
"The first time you run a program under Acme::DonMartin, your source code is magically transformed into Don Martin cartoon sound effects. [...] This can also be construed as a security feature. It is expected that a hacker will be laughing too hard to be able to recover the source code."
Nice idea, but for that to work the router on the receiving end of the connection has to know how it should route the encapsulated packet differently from a normal packet, and the sending end would need to know the 'non-routeable but addressable' IP address of the destination - that is, assuming you want to be able to address each computer in the network individually, which is somewhat important to your point; if it wasn't, you'd just be reinventing NAT with a larger unrouteable address space.
The big upside of your proposal is that on the router side of things, only the router at the destination needs to know how to extract the inner packet and route it. It's also not hard at all to have a big chunk of IP addresses; someone could have a nice/32/16 block of 65536 different and addressable IP addresses without the Internet blinking an eyelid.
The downsides, however, are quite a lot. IPv4-in-IPv4 would necessarily dictate that the outer packet is sent to a single IP address. If that IP address gets DDoSed or goes down in some other way, you lose the entire network behind that router. That could amount to a lot of machines, especially if the encapsulation was used as an excuse to begin the "plug everything into the Net!" drive.
The other downside - and one that couldn't be ignored - is that any software wanting to use this technology *has* to be aware of it. Current IPv4 apps will only be able to access a maximum of 256^4 addresses. You could probably write a driver that lets you 'page' to another IP block in much the same way as EMS worked for memory, but there would be no way to have that program automatically page for you, as there wouldn't be any way to tell which IP block you wanted (and even if it could, you couldn't specify both the block *and* the secondary IP address you want to reach at the same time).
It's a nice idea, and it could even be implemented on a very small scale. But IPv6 already has the fame it needs; programs are aware of it, and there are already people offering IPv6 addresses, whether through tunnel brokers or native connections. Plus, IPv6 is more than just a huge address space; it also offers native IPSec functionality, and IPv6 was designed with ease of routing in mind.
Like it or not, IPv6 is the future. We can't go back on it now, and to be honest, I don't think we'd want to either.
...I don't mean to say that I necessarily agree with this guy, but it seems to me that this post shouldn't have been modded down. It's a viewpoint commonly held (and perhaps erroneously, but that's not the point here) by many people. He isn't a troll because he probably didn't post just to cause an argument. IMO, anyway.
As for my own opinion, I don't know. I haven't read up enough to make an informed decision about it. However, since this is about free software, and the whole reason Africa is in poverty is because of the debt owed the country, this can only be a good thing.
You're probably using a localised Google homepage. It only appears on the main Google.com page - click on the "Go to Google.com" link in your localised page to see it.
I use a localised page too, so I didn't see it at first either.
Seriously, I think anybody who buys the software will want to do so with their own setup and everything. You sound like you're suggesting either a LiveCD, or a fully-installable distro. Both have their problems; most notably, you can't use the user's own home directory because it won't have everything you want in exactly the right setup. Secondly, installation of a distro takes time. It may be quicker than Windows but that's not to say you can't still spend a lot of time doing it.
Using your own distro in this way would presumably mean you use your own package format - otherwise why not just release in a normal RPM, deb, or whatever? Most distros have (and can use) rpm or apt.
Making a distro takes time in itself, too, and there is absolutely no way you're going to please anybody with one particular distro. Not to mention that if this is a big enterprise thing, they might have to install the program over lots of computers, hence the *need* for the easy installation. Or maybe it's designed to go onto a server, in which case the disadvantages of having to install a distro or use a LiveCD on a working, production server that's already set up should be obvious.
Besides, installing a distro (whether it's your own or not) just for one program is very ego-centric.
The link only shows up if you're in the US. I know, because I couldn't see it either, but when I switched to using a proxy in the US and then refreshed, the link showed up.
And that's why you don't even have one, right? Oh, sorry, you don't want to risk your precious karma. Meh.
For the record, I agree that using Apache (or any Web server) at home isn't the same. There are all sorts of complications that go along with that - the most obvious being low bandwidth and uncertain uptime. And if your site gets Slashdotted on a home connection... well, good luck to you when your ISP finds out. You'll need it.
Unlike some other people here, though, I don't go about saying that it is "the stupidst post ever". There are plenty of posts that are more stupid than that. (No, I'm not giving into the temptation of saying the obvious next line, because it isn't true.)
Think a bit further. These are spam blogs. No one in their right mind is actually going to read them, so there must be another reason for their existence - and there is.
Search engines will take the links and rank them up. Not good.
I'm not chewing you out over this, but I really am curious; how long ago was this? Because nVidia drivers not only exist nowadays, but they're pretty darn good. (If you discount their closed-sourcedness, but then that's understandable to an extent.)
nVidia and ATI are the two major players in Linux graphics card drivers, and nVidia have done a lot of work for Linux.
Slashdot Mirror
I believe that disclosing how your site is performing in that way is against the Google AdSense Terms and Conditions (see clause 7). You can give the total amount of payment, but you've also given the number of clicks you get, which the T&Cs disallow.
I believe the implication was that the parent believes that the large number of stories about Google are some sort of advertising method.
Disclaimer: I wasn't able to watch the video for some reason. I only heard audio. So I can't tell how obvious it actually is.
So you're saying that you wanted Google to hand over its search data?
1) The statement made was false.
2) The statement made was fact, not opinion.
Don't these two things contradict each other? Did yoiu mean "The statement made was presented as fact, not opinion"?
Since then, they initiated a program where people can buy their web site rankings.
What program is that?
That link was about AdSense. That was ages ago.
This is about ads next to Google's own search results.
Did you see the date on that entry?
"Posted: Mon - May 17, 2004 at 03:39 PM"
That was talking about AdSense. *This* is talking about Google itself.
I'm sure Rogers would love to see your foolproof solution. I mean, you obviously have one.
Seriously, I agree with your reasoning for most things. And yes, this definitely is security through obscurity, which is a bad thing. However, how else would you do an automatic fraud detection system on phone calls? Of course, you could always give the (probably correct) response that automatic fraud detection is a bad thing as it's never going to be perfect, and there will always be those who slip through, as in this case.
Much as I hate to admit it, though, I don't know how else it could be implemented. Fraud is a serious thing that needs to be watched for, and there would be an outcry if Rogers removed it.
It's p2p, yes. But p2p needs some servers to search with. Assuming they can get their new software onto the SuperNodes (if they use different terminology now, please correct me), the server would be able to block searches no problem. It might be hit-and-miss sometimes, though.
Exploiting a flaw in this decoding is as simple as exploiting any other software bug.
Perhaps - except that according to the Wikipedia entry on the Red Book standard, it's stored in PCM format, which to my knowledge can be passed straight to the soundcard with no further processing. The only decoding in the procedure is done by the CD drive and the soundcard, and while exploiting a flaw in either of those might not be difficult, it would also be highly dependent on the chipset used, and in the end would probably both not be able to do much in the grand scheme of things, *and* would probably cause other CD drives/soundcards to barf.
Acme::DonMartin is the way to go.
"The first time you run a program under Acme::DonMartin, your source code is magically transformed into Don Martin cartoon sound effects. [...] This can also be construed as a security feature. It is expected that a hacker will be laughing too hard to be able to recover the source code."
So, let me get this straight.
These guys want to measure how often people listen to podcasts.
Podcasts. As in, things you listen to on an iPod, or other mobile music player.
Can these things even connect to the Internet at the moment?
Nice idea, but for that to work the router on the receiving end of the connection has to know how it should route the encapsulated packet differently from a normal packet, and the sending end would need to know the 'non-routeable but addressable' IP address of the destination - that is, assuming you want to be able to address each computer in the network individually, which is somewhat important to your point; if it wasn't, you'd just be reinventing NAT with a larger unrouteable address space.
/32/16 block of 65536 different and addressable IP addresses without the Internet blinking an eyelid.
The big upside of your proposal is that on the router side of things, only the router at the destination needs to know how to extract the inner packet and route it. It's also not hard at all to have a big chunk of IP addresses; someone could have a nice
The downsides, however, are quite a lot. IPv4-in-IPv4 would necessarily dictate that the outer packet is sent to a single IP address. If that IP address gets DDoSed or goes down in some other way, you lose the entire network behind that router. That could amount to a lot of machines, especially if the encapsulation was used as an excuse to begin the "plug everything into the Net!" drive.
The other downside - and one that couldn't be ignored - is that any software wanting to use this technology *has* to be aware of it. Current IPv4 apps will only be able to access a maximum of 256^4 addresses. You could probably write a driver that lets you 'page' to another IP block in much the same way as EMS worked for memory, but there would be no way to have that program automatically page for you, as there wouldn't be any way to tell which IP block you wanted (and even if it could, you couldn't specify both the block *and* the secondary IP address you want to reach at the same time).
It's a nice idea, and it could even be implemented on a very small scale. But IPv6 already has the fame it needs; programs are aware of it, and there are already people offering IPv6 addresses, whether through tunnel brokers or native connections. Plus, IPv6 is more than just a huge address space; it also offers native IPSec functionality, and IPv6 was designed with ease of routing in mind.
Like it or not, IPv6 is the future. We can't go back on it now, and to be honest, I don't think we'd want to either.
...I don't mean to say that I necessarily agree with this guy, but it seems to me that this post shouldn't have been modded down. It's a viewpoint commonly held (and perhaps erroneously, but that's not the point here) by many people. He isn't a troll because he probably didn't post just to cause an argument. IMO, anyway.
As for my own opinion, I don't know. I haven't read up enough to make an informed decision about it. However, since this is about free software, and the whole reason Africa is in poverty is because of the debt owed the country, this can only be a good thing.
You're probably using a localised Google homepage. It only appears on the main Google.com page - click on the "Go to Google.com" link in your localised page to see it.
I use a localised page too, so I didn't see it at first either.
Hey, good idea! That's really clever!
Seriously, I think anybody who buys the software will want to do so with their own setup and everything. You sound like you're suggesting either a LiveCD, or a fully-installable distro. Both have their problems; most notably, you can't use the user's own home directory because it won't have everything you want in exactly the right setup. Secondly, installation of a distro takes time. It may be quicker than Windows but that's not to say you can't still spend a lot of time doing it.
Using your own distro in this way would presumably mean you use your own package format - otherwise why not just release in a normal RPM, deb, or whatever? Most distros have (and can use) rpm or apt.
Making a distro takes time in itself, too, and there is absolutely no way you're going to please anybody with one particular distro. Not to mention that if this is a big enterprise thing, they might have to install the program over lots of computers, hence the *need* for the easy installation. Or maybe it's designed to go onto a server, in which case the disadvantages of having to install a distro or use a LiveCD on a working, production server that's already set up should be obvious.
Besides, installing a distro (whether it's your own or not) just for one program is very ego-centric.
If you're using Firefox, you can select how you want applications to work with new windows in the Advanced section of the Options/Preferences window.
I don't know if there's a similar thing for IE/Opera/any other browser, though.
The link only shows up if you're in the US. I know, because I couldn't see it either, but when I switched to using a proxy in the US and then refreshed, the link showed up.
Actually, it does if you send it the start of a valid Jabber stream. Take a look at the other comments to the parent.
Um, patents do exist in Europe. I think you're getting confused with software patents.
Besides which, that idea is plainly unpatentable.
(yes, I know I shouldn't feed the trolls. Sorry.)
And that's why you don't even have one, right? Oh, sorry, you don't want to risk your precious karma. Meh.
For the record, I agree that using Apache (or any Web server) at home isn't the same. There are all sorts of complications that go along with that - the most obvious being low bandwidth and uncertain uptime. And if your site gets Slashdotted on a home connection... well, good luck to you when your ISP finds out. You'll need it.
Unlike some other people here, though, I don't go about saying that it is "the stupidst post ever". There are plenty of posts that are more stupid than that. (No, I'm not giving into the temptation of saying the obvious next line, because it isn't true.)
Think a bit further. These are spam blogs. No one in their right mind is actually going to read them, so there must be another reason for their existence - and there is.
Search engines will take the links and rank them up. Not good.
Right. And according to the history, it never linked to Penis either.
I thought mods would at least look at what they were modding down.
I'm not chewing you out over this, but I really am curious; how long ago was this? Because nVidia drivers not only exist nowadays, but they're pretty darn good. (If you discount their closed-sourcedness, but then that's understandable to an extent.)
nVidia and ATI are the two major players in Linux graphics card drivers, and nVidia have done a lot of work for Linux.