Precisely. I read the patent, and it describes *exactly* the operation of a 556 timer based PC game port. The only unique thing that I can spot in this patent is listed on the first page - "This implementation provides a joystick port which uses low-voltage CMOS VLSI structures..."
So to get definite prior art, you just need to find a soundcard with a standard 0x201 gameport address, with the game port hardware implemented in a CMOS ASIC. Which is really easy, considering the patent was filed for in 1998. Even in the early-mid 90's, cheap "AOpen to Zoltrix" soundcards consisted of a single ASIC which handled the ISA or PCI interface, A/D+D/A conversion, 0x201 game port, MIDI interface and the works. And there's a 100% chance that said ASIC is a CMOS process.
What I find silly is that they're suing Microsoft, Sony and Nintendo over this... I seriously doubt that any of their hardware uses a timer approach for digitizing analog inputs - I'm willing to bet they use true ADC's. The 556 timer hack was useful in the 80's PC days when silicon was expensive and discretes and PCB space were cheap, but today when a $2 PIC microcontroller can have almost a dozen 10+ bit ADC inputs, there's no point anymore.
Since most infected computers on this botnet are XP SP2 and likely have Windows Firewall enabled on them...
How hard can it be for MS to code up a patch to the firewall code that detects outgoing connections to TCP port 25 (SMTP) and throws a warning on the screen? Send the patch out over Windows Update.
Your average Hotmail/Yahoo/Gmail user won't ever notice. People who use Outlook Express or some other SMTP-sending client may have to click a "yes, I'm actually sending e-mail" button when they send e-mail and suffer half a second of annoyance, and that's just assuming you alarm on every outgoing SMTP connection. There are probably better ways to do it.
Something like this would completely wreck SpamThru's functionality, wouldn't it?
Just a thought.
All the stuff that Creative's marketing department spews about MP3's sounding better than the original CD could be "right"... Sound is subjective, and for all I know the X-Fi probably applies bass boost, treble boost and mid cut similar to the "ROCK" equalizer preset that can be found on any portable MP3 player or car stereo. Which for lots of people, makes music sound better.
(of course, this doesn't mean it's not a load of crap. There's still that whole 'information theory' thing to deal with...)
But anyway, taking a 44.1/16 file and upsampling it to 96KHz, keeping it at 24 bits after upsampling instead of dithering/truncating to 16 bits, *then* driving a DAC with the resulting 24/96 signal will give you excellent sound quality. Assuming of course that you have a good interpolation filter - since the X-Fi is an audio processing monster, there's no reason they can't do a good job of it.
Lots of CD players have "4x oversampling DAC" stickers on the front - this is the same thing. The incoming 44.1 audio is interpolated 2 or 4 times using a digital filter chip, and the resulting 88.2 or 176.4KHz audio stream is then fed to a pair of DAC channels. The result is less distortion, less high frequency garbage and less mangling of the audio signal by the analog filter that follows the DAC. Which is a good thing.
And the X-Fi has some good things - while most common wavetable chips which do linear or simpson interpolation to do a sample rate change, each "voice" on the X-Fi uses an ASRC (asynchronous sample rate converter) to perform it with far less distortion. Which means with a good patch set, it will be a very good sounding MIDI card.
My only complaint about the card is that I'll never be able to get my hands on the specs to program it... Oh, and I probably won't buy one because I'm still bitter about Aureal.:)
... so I can't really see why people are recommending bluetooth, 802.11, etc. I'm convinced that you guys are all computer science graduates, not engineers that develop real products;)
What this guy needs is a simple system - a 1-way RF transmitter with a couple of buttons on it. Just like a garage door opener, a keyless remote for a car, etc. Remember that these things are dirt simple, have excellent range, are individually serialized, and are extremely cheap to make.
You could probably build the remotes yourself - you're basically talking something like an rfPIC from Microchip with a couple of buttons on it. I think Philips, Cypress, etc. also make similar solutions... But you could probably buy a box of 500 car remotes from China for less than the price of making 10 of your own remotes.
For the price, you probably won't even care if students occasionally walk off with the remotes. Or heck, you might even be able to let them - eg. give a student their own remote for a semester, which reduces the "which RF transmitter ID matches which student" problem to a once-a-semester thing. A student could possibly also use a single remote for multiple classes.
Then it's a matter of finding a receiver. You can probably buy an evaluation kit from the car remote manufacturer, which includes a few remotes and a receiver which plugs into a computer. Then it's all software from that point on.
I'd do a bit of searching in this area. Hope this helps.
Slashdot Mirror
It's in every second post right now, but yes, get this book.
Precisely. I read the patent, and it describes *exactly* the operation of a 556 timer based PC game port. The only unique thing that I can spot in this patent is listed on the first page - "This implementation provides a joystick port which uses low-voltage CMOS VLSI structures..."
So to get definite prior art, you just need to find a soundcard with a standard 0x201 gameport address, with the game port hardware implemented in a CMOS ASIC. Which is really easy, considering the patent was filed for in 1998. Even in the early-mid 90's, cheap "AOpen to Zoltrix" soundcards consisted of a single ASIC which handled the ISA or PCI interface, A/D+D/A conversion, 0x201 game port, MIDI interface and the works. And there's a 100% chance that said ASIC is a CMOS process.
What I find silly is that they're suing Microsoft, Sony and Nintendo over this... I seriously doubt that any of their hardware uses a timer approach for digitizing analog inputs - I'm willing to bet they use true ADC's. The 556 timer hack was useful in the 80's PC days when silicon was expensive and discretes and PCB space were cheap, but today when a $2 PIC microcontroller can have almost a dozen 10+ bit ADC inputs, there's no point anymore.
Since most infected computers on this botnet are XP SP2 and likely have Windows Firewall enabled on them... How hard can it be for MS to code up a patch to the firewall code that detects outgoing connections to TCP port 25 (SMTP) and throws a warning on the screen? Send the patch out over Windows Update. Your average Hotmail/Yahoo/Gmail user won't ever notice. People who use Outlook Express or some other SMTP-sending client may have to click a "yes, I'm actually sending e-mail" button when they send e-mail and suffer half a second of annoyance, and that's just assuming you alarm on every outgoing SMTP connection. There are probably better ways to do it. Something like this would completely wreck SpamThru's functionality, wouldn't it? Just a thought.
All the stuff that Creative's marketing department spews about MP3's sounding better than the original CD could be "right"... Sound is subjective, and for all I know the X-Fi probably applies bass boost, treble boost and mid cut similar to the "ROCK" equalizer preset that can be found on any portable MP3 player or car stereo. Which for lots of people, makes music sound better.
:)
(of course, this doesn't mean it's not a load of crap. There's still that whole 'information theory' thing to deal with...)
But anyway, taking a 44.1/16 file and upsampling it to 96KHz, keeping it at 24 bits after upsampling instead of dithering/truncating to 16 bits, *then* driving a DAC with the resulting 24/96 signal will give you excellent sound quality. Assuming of course that you have a good interpolation filter - since the X-Fi is an audio processing monster, there's no reason they can't do a good job of it.
Lots of CD players have "4x oversampling DAC" stickers on the front - this is the same thing. The incoming 44.1 audio is interpolated 2 or 4 times using a digital filter chip, and the resulting 88.2 or 176.4KHz audio stream is then fed to a pair of DAC channels. The result is less distortion, less high frequency garbage and less mangling of the audio signal by the analog filter that follows the DAC. Which is a good thing.
And the X-Fi has some good things - while most common wavetable chips which do linear or simpson interpolation to do a sample rate change, each "voice" on the X-Fi uses an ASRC (asynchronous sample rate converter) to perform it with far less distortion. Which means with a good patch set, it will be a very good sounding MIDI card.
My only complaint about the card is that I'll never be able to get my hands on the specs to program it... Oh, and I probably won't buy one because I'm still bitter about Aureal.
... so I can't really see why people are recommending bluetooth, 802.11, etc. I'm convinced that you guys are all computer science graduates, not engineers that develop real products ;)
What this guy needs is a simple system - a 1-way RF transmitter with a couple of buttons on it. Just like a garage door opener, a keyless remote for a car, etc. Remember that these things are dirt simple, have excellent range, are individually serialized, and are extremely cheap to make.
You could probably build the remotes yourself - you're basically talking something like an rfPIC from Microchip with a couple of buttons on it. I think Philips, Cypress, etc. also make similar solutions... But you could probably buy a box of 500 car remotes from China for less than the price of making 10 of your own remotes.
For the price, you probably won't even care if students occasionally walk off with the remotes. Or heck, you might even be able to let them - eg. give a student their own remote for a semester, which reduces the "which RF transmitter ID matches which student" problem to a once-a-semester thing. A student could possibly also use a single remote for multiple classes.
Then it's a matter of finding a receiver. You can probably buy an evaluation kit from the car remote manufacturer, which includes a few remotes and a receiver which plugs into a computer. Then it's all software from that point on.
I'd do a bit of searching in this area. Hope this helps.