It's easy to spoof, but it's hard to establish a TCP connection by only sending outgoing packets. (Spoofing is the most useful for attacks on UDP protocols.)
I don't think that spammers have any amazing tech, they just have different requirements. They can still send spam with a 1% success rate whereas with OCR you'd want a 99% success rate.
When I moved out of my mother's basement I used the Albertson's delivery service until they shut it down. It was $14 per delivery, regardless of size, so I'd get all of my groceries for the month in one order.
It was a lot easier to avoid impulse buying and to plan out what was actually needed when I could place the order online. Albertson's would remember your previous order so it was easy to just adjust it slightly each month.
I imagine that google is just going to restrict the use of the android trademark, not source code. This seems like a sensible solution, and won't restrict the rights of the manufacturers.
Also check out Musopen, a large collection of public-domain classical music recordings and sheet music. They take donations and use those donations to hire professional artists to make new recordings of the pieces and then put them into the public domain.
I think you're confusing speech recognition with natural language parsing. They are two different components. The reason natural language processing is powerful is the same reason the command line is powerful. With a GUI, you normally have to find your way through menus to get to a particular functionality. When there aren't many options to present or you aren't familiar with the system, this is a good interface. On the other hand, once you're familiar, a multitude of functionality is only a command away. (I have 4503 commands in $PATH.)
Once the speech recognition reaches a certain point you'll be able to call your bank and say something like "transfer $500 from checking to savings".
Anyway, here's the use case I picture: Anywhere in the house, I can say something in a normal voice, like "computer, what's the best way to stop a bloody nose?" or "computer, how long do i need to boil an egg?" or "computer, turn on the front-porch lights". Being able to interact with a phone in the same manner would be appropriate in some situations as well. Speech has the potential to be a great input device, especially when there isn't a keyboard handy. (Insert joke here about how none of us get far enough away from a keyboard for it to matter.)
What's the difference between this and the SMS service already offered by Facebook? Facebook's service works even if you don't have a special SIM card, and the last time I checked (several years ago) it had lots of useful features.
This is a great cause, and I'm sure it will help a lot, but it won't completely eliminate spam. If your contacts are phished or are running malware then you will start receiving mail from your trusted contacts.
A good example of this happening right now is the spam that happens on facebook. The only people allowed to write on my wall are people that I personally know (just one step away in the ring of trust) and yet I have still received spam.
Not to mention, unless I'm much mistaken a NAT can support 65536 connections at maximum (number of valid ports for outgoing connections).
I believe that with TCP the limit is 65536 concurrent connections to one host and port. In other words, you can use the same source port for two concurrent, outgoing connections at the same time as long as the destination is different. (This is similar to the principle of how more than one connection to a single port on a server work without problems as long as the client host/ip combination is different.)
If you have a fast CPU (which I imagine most people who are trying to encode HD content will have) then you could possibly compress it into an intermediate format on the fly, reducing the I/O requirements to what could be handled by a single spinning hard drive.
There are thousands of files that get openned when an app like Firefox for Photoshop starts running, and all of those files get read into memory and checked... BEFORE they are passed along to the app calling them. Unless you invent time bending or something, this will always end up taking a very noticeable amount of time, making your computer seem slow.
This isn't time bending, but it might be mind bending. Use a kernel hook to check files as they are written and you don't have to do anything when they are read.
From there you can get more sophisticated and queue up files that are written and process that queue as an idle task while the data is still in memory cache. A simple read hook could do immediate scanning of files that are requested before they have been processed by the idle job.
Slashdot Mirror
Why not just put a QR code near the entrance of the museum, or a web address?
Coral Cache isn't working for me, but google has cached it.
Someone should host the PDF on S3.
You could create a lot of variants if you just randomly drop a word from each one.
People have always been smart, stupid is the exception.
I would estimate that roughly half of people have less-than-average intelligence.
It's easy to spoof, but it's hard to establish a TCP connection by only sending outgoing packets. (Spoofing is the most useful for attacks on UDP protocols.)
I don't think that spammers have any amazing tech, they just have different requirements. They can still send spam with a 1% success rate whereas with OCR you'd want a 99% success rate.
This seems like an amateur mistake. Who are these companies hiring lately?
At the very least, hash and salt. If the hashes might be stolen then hash it thousands of times (see PBKDF2).
When I moved out of my mother's basement I used the Albertson's delivery service until they shut it down. It was $14 per delivery, regardless of size, so I'd get all of my groceries for the month in one order.
It was a lot easier to avoid impulse buying and to plan out what was actually needed when I could place the order online. Albertson's would remember your previous order so it was easy to just adjust it slightly each month.
I imagine that google is just going to restrict the use of the android trademark, not source code. This seems like a sensible solution, and won't restrict the rights of the manufacturers.
Also check out Musopen, a large collection of public-domain classical music recordings and sheet music. They take donations and use those donations to hire professional artists to make new recordings of the pieces and then put them into the public domain.
I think you're confusing speech recognition with natural language parsing. They are two different components. The reason natural language processing is powerful is the same reason the command line is powerful. With a GUI, you normally have to find your way through menus to get to a particular functionality. When there aren't many options to present or you aren't familiar with the system, this is a good interface. On the other hand, once you're familiar, a multitude of functionality is only a command away. (I have 4503 commands in $PATH.)
Once the speech recognition reaches a certain point you'll be able to call your bank and say something like "transfer $500 from checking to savings".
Anyway, here's the use case I picture: Anywhere in the house, I can say something in a normal voice, like "computer, what's the best way to stop a bloody nose?" or "computer, how long do i need to boil an egg?" or "computer, turn on the front-porch lights". Being able to interact with a phone in the same manner would be appropriate in some situations as well. Speech has the potential to be a great input device, especially when there isn't a keyboard handy. (Insert joke here about how none of us get far enough away from a keyboard for it to matter.)
What's the difference between this and the SMS service already offered by Facebook? Facebook's service works even if you don't have a special SIM card, and the last time I checked (several years ago) it had lots of useful features.
I believe that it's just WHATWG that has dropped the version number, not the W3C.
Also note that 32-bit operating systems can still make use of larger system memory sizes.
I'm doing this right now. It's as easy as apt-get install linux-image-generic-pae.
Didn't they launch chrome just a few years ago? I haven't read the summary yet, but this sure is a shame.
This is a great cause, and I'm sure it will help a lot, but it won't completely eliminate spam. If your contacts are phished or are running malware then you will start receiving mail from your trusted contacts.
A good example of this happening right now is the spam that happens on facebook. The only people allowed to write on my wall are people that I personally know (just one step away in the ring of trust) and yet I have still received spam.
When you want to slow down a fast hash you just do it a lot of times. See PBKDF2, for example.
I've heard of marketers redefining price points, but this is ridiculous. I've never paid more than $150 for a processor.
Not to mention, unless I'm much mistaken a NAT can support 65536 connections at maximum (number of valid ports for outgoing connections).
I believe that with TCP the limit is 65536 concurrent connections to one host and port. In other words, you can use the same source port for two concurrent, outgoing connections at the same time as long as the destination is different. (This is similar to the principle of how more than one connection to a single port on a server work without problems as long as the client host/ip combination is different.)
And even if that ever fails, someone can always put up a version on TPB.
Actually I think facebook has you load their javascript code asynchronously, which would avoid the issue you mention:
http://developers.facebook.com/docs/reference/javascript/
If you have a fast CPU (which I imagine most people who are trying to encode HD content will have) then you could possibly compress it into an intermediate format on the fly, reducing the I/O requirements to what could be handled by a single spinning hard drive.
PAE
There are thousands of files that get openned when an app like Firefox for Photoshop starts running, and all of those files get read into memory and checked ... BEFORE they are passed along to the app calling them. Unless you invent time bending or something, this will always end up taking a very noticeable amount of time, making your computer seem slow.
This isn't time bending, but it might be mind bending. Use a kernel hook to check files as they are written and you don't have to do anything when they are read.
From there you can get more sophisticated and queue up files that are written and process that queue as an idle task while the data is still in memory cache. A simple read hook could do immediate scanning of files that are requested before they have been processed by the idle job.
This idea was invented by Shampoo.