Honestly, how long can it possibly take to sort 300 shows? It's just three hundred! I have a crappy java program here at work that sorts lists of thousands faster than you can say "I hate my tivo."
I would expect 300 shows to take longer than 20, but I'd expect 20 to take about 1/10 of a second. What can possibly be taking so long for the tivo to do?
Another question-- why are some people's boxes worse than others? Mine is slow, yours is not as slow, another poster says he's got no issue at all. A good friend of mine can't reorder his passes at all, except overnight, because it literally takes several hours to do. Why the variance?
Interesting... I wonder what the variable is? I have a number of friends mad at me for selling them on the DirecTivo based on my Series 1 standalone tivo experience-- they're all locked into year-long contracts and having the same troubles I am, some worse. There's clearly a set of conditions that need to be met if you're not having the problems-- now we just need to figure out why yours works, and mine and my friends' don't.
I'll see if anybody's got aftermarket RAM for the units-- but I think it's soldered on the motherboard. And I'll go back to my original question-- how much RAM does it take to sort a list of 30 entries? I think my C64 could handle it nearly instantaneously. What is the tivo doing that takes more than an eyeblink?
Honestly, it sounds like a job for a first-year CS student. DirecTV has really bungled the Tivo software somehow. How long can it possibly take to reorganize a list of 30 entries? Or to read the guide database and print out 10 entries at a time? And if the backend stuff takes so long, why doesn't it get done in the background while the user goes back to watching TV? Even if there is a good reason that reordering your season passes takes 10 minutes, the UI should return control to the user and chug away in the background.
There's a 200MHz CPU in there, if I remember right. Why is it sooooo slow?
He needs to get into the DirecTV DVR code and figure out why it takes 30 seconds to display the guide, a minute to open your "Now Playing" list of shows, and 5+ minutes to sort a 30-entry list of season passes.
A huge fraction of Tivo's subscriber base is through the DirecTV tivos-- and despite my great experience with the standalone unit I had, the DirecTV box is so much slower despite 4x the processor speed that I can't even imagine what sort of horrible code is in there. Optimize the UI, *then* add features. DirecTV may singlehandedly turn millions of people away from tivo after they sign up and have a truly subpar experience with it.
The more primaries, the merrier. The gamut gets bigger the more you stick in there. A nice six-primary standard for displays would make me very happy... but so would a pocket-sized two-CCD digicam that could take 3D pictures. We can all dream.
As to Velvia, I understand it has a wider gamut than other film, but I haven't the faintest idea why. Perhaps the chemicals (and thus the color response) they've chosen correspond more closely to the response of our eyes? I'd be interested to know, if you've come across any information on what exactly makes Velvia so good.
My god I loved that game, and the spectacular sequel. Fighting ninjas in a trailer park while a tornado carries your trailer into the sky? Exploding robot kitten anti-personnel mines? Schoolgirl-aged assasins talking about going shopping before they try to chop your head off? A getaway sequence on a tricycle shooting evil mimes? Man-cubes as furniture? It goes on and on. I have never laughed so hard in my life. I cannot recommend NOLF and NOLF 2 enough.
You're certainly right about it being non-obvious. It took me a while to wrap my head around it, too.
The problem isn't that they've chosen the wrong primaries, although you're right that a different set could cover better than the ones we're using. The problem is that the three types of color sensors in the eye are not perfect single-frequency sensors-- they have wide, irregular frequency responses that overlap significantly. There are some spots where all three receptors would fire from a single-frequency beam of light, notably around 450-475, looking at the chart. How can you represent a different blue with a blue in that part of the spectrum if it fires the red (and/or green) receptors? It will inevitably look purple. The "negative" amount of red is just a mathematical artifact, a way to say "to represent this blue, you'd need a negative amount of red to cancel the phantom red caused by the blue light"-- and you simply can't do that in real life. As a photographer, you may have noticed effects like this occasionally in photographs. I have a shot of a beautiful deep blue flower I took in Arizona that is a crappy purple color in every shot I took. The film or CCD (both are RGB) you use to capture the image will produce the occasional noticable color distortion.
It's because of this overlap that we can't pick just three primaries that cover all the colors we can see. If the peaks were non-overlapping, it would work much, much better. So we need more complex models with nonlinear color mapping functions like CIE XYZ or CIE L*a*b to fully exploit the potential of the overlapping wide-response receptors in our eyes.
It is difficult to show with nothing but an RGB computer monitor, since it can't display the colors it can't display, but you might be able to find a Pantone book somewhere and look up some colors that are out of gamut for RGB displays, scan them or take a picture, and then look at the difference on your screen. You've probably noticed a similar effect when you try to print color images, as well, since color printing is CMYK, and not all RGB colors are in the gamut for CMYK.
D'oh! Well, scratch that part from what I said. I was unaware that CIE had used such a limited dataset in establishing the gamut. In my opinion, this is the sort of thing the HDTV standards board should have done-- a wide study to establish an accurate colorspace and delivery standard, so that we could have a nice, full-gamut (or as full as practical) display standard to go with our shiny new high-resolution televisions.
RGB doesn't cover the visible gamut. At all.
on
RGB to become RGBCMY
·
· Score: 2, Insightful
It's not the discrete gaps that are the problem! RGB does not represent all of the visible colors, even theoretically. Assuming a perfectly smooth RGB model with infinite intensity and perfect black, and infinitely precise levels of R, G, and B, there is a huge chunk (around 45%, if I remember right) of the visible gamut that is totally unreproducible. CMY covers some areas that RGB doesn't, and vice versa. Neither is the whole gamut. There are more complex models that do, like CIE L*a*b.
Re:RGBCMY is more marketing factoid than it isreal
on
RGB to become RGBCMY
·
· Score: 1
Not quite. The "step" issue is separate from the fact that the RGB gamut does not cover the visible gamut all the way. There are colors we can see that no amount of twiddling can *ever* get an RGB monitor to reproduce.
You are right that a digital RGB representation is discrete, not smooth, but there are colors "outside the grid," too. Pure yellow, for example.
Not so. RGB, CMY, YUV, etc... are not full gamut.
on
RGB to become RGBCMY
·
· Score: 3, Informative
RGB, CMY, CMYK, etc... *cannot* represent the entire visible color gamut. YIQ (the one used by NTSC TV), YUV (PAL TV), and YCrCb represent a smaller gamut than RGB, to be sure, but neither represent the whole thing.
For that, you need a more complex model like CIELAB.
Quake 3 had used a physics model that did the calculations per-frame on the client side. To save CPU, this stuff isn't terribly precise, so there's some rounding error. Because of that, there were some "ideal framerates" where the rounding errors added up in your favor, resulting in the ability jump farther or run faster-- which is why the true nutcases built systems that could run at a framerate never below 125 (or whatever they believed/calculated the max to be), and then *locked* the framerate there.
http://ucguides.savagehelp.com/Quake3/FAQFPSJump s. html
has more info, along with a chart of where the best spots to lock your framerate are.
I'm not 100% sure, but I vaguely remember there being talk of a 60fps framerate lock in Doom III to prevent this sort of cheating.
People in every game work to get high framerates for other reasons, though-- really high *average* framerate means fairly high *minimum* framerate. And of course, minimums come right when you need it to be smooth-- like when 16 guys in jeeps, tanks, and airplanes all come down on you at once in BF1942.
It may seem absurd, but there are legitimate reasons why 3D cards need a higher framerate to represent the same smooth motion of 24fps movies.
I've explained this before, and I'll do it again. Television and video have motion blur-- the effect of the capture device essentially "averaging" the motion that occurred across the duration of capturing that frame.
Video cards generate a crisp, instantaneous frame that represents only the precise instant the frame was rendered, not the whole time the "shutter" was open.
At a *bare minimum* producing motion that looks as smooth as blurred 24fps requires double that. (You have to have two frames for your eye to blur between) To do it as well as a film camera requires even more, since their motion blur is effectively an infinite number of samples averaged together over the duration the shutter was open. I'd guess you could get a reasonable approximation at 3x the framerate.
TV and Movies are also filmed with the 24fps limitation in mind-- good cinematographers are well aware of the limits and know how to avoid situations that would result in jerky movement.
Tivos are a mixed bag. Series 2 units are fast and responsive compared to the Series 1 units, but DirecTV tivos (which are basically Series 2 units) are so slow that it can take 30 seconds to re-order your season passes, 5 seconds for the "now playing" list to pop up, etc...
Why you can't display or re-order a list of 30 text entries in less than a second with a 200MHz CPU baffles me. Sure, it won't play Doom III, but it really ought to respond more quickly to UI interactions. I'm not sure why the DirecTV Tivos are so awful-- it's the same hardware as the Series 2 Standalones.
Nicely done. I salute your metaphor-salvaging effort! But as the further nitpicking efforts by other posters point out, it probably ought to read something like:
"Selling Cooling Machinery to the Indiginous Peoples Who Populate Areas Where It Is Always Colder Than The Lowest Temperature Obtainable Using The Machinery."
Which isn't as cool. And if it's really 54 where the Inuit are, and you can actually sell them Igloos, they might be interested in your special frozen-while-it's-above-freezing Igloo, which would make your original metaphor valid without any fudging.
I should just stop nitpicking. Everybody knew what you meant.:)
I believe you meant "selling iceboxes to eskimos." They might actually have a use for an igloo, as long as you can show that yours are a superior value over the homebuilt variety.
I should have posted longer, but was trying to keep it short. A plain old dremel worked almost as quickly, but needed two discs. We played around with it for quite a while after the first cut-- there's very little those locks protect against.
I don't think you'd have an issue carrying a battery-powered dremel and a dozen cutting discs to wherever you wanted to cut the lock.
Bolt cutters couldn't do it, to be sure. Abrasive cutting discs, no problemo.
Those "super-tough" Kryptonite U-locks take all of 15 seconds to cut through. I had a friend lock her wheel to the frame on her bike, and lose the key in a move. We were all set to resort to drastic measures with a blowtorch or liquid nitrogen, but all it took was compressed-air cutting tool (like a dremel, but powered like an impact wrench) with a proper cutting disc. Radiac, if I remember right. Went through it like butter. I've kinda lost all my faith in security products since then-- the first thing we tried went through it before you could say "hey, where's my bike?"
The shower of sparks might be a giveaway in a public place, though, so at least you've got that going for you.
Holy crap, somebody beat us to it! Ah well... I suppose the lesson is "no matter how dorky you you are, there's always someone dorkier."
Kudos to the folks setting this up-- it may be pointless, awkward to play, and a pain to set up, but that hasn't stopped me from wasting thousands of hours on similar futile excercises in geekery in the past, either.
You're right-- it's the same damn game. "Memorize sequence of arbitrary steps/directions and enter them on the pad/joystick at exactly the right moment.
You could probably hook the dance pad right up to an old DL machine, and play the game that way. If there's a PC version of DL, I'll bet you could use a PC DDR pad with it out-of-the-box. Of course, the steps are farther apart, if my old amiga version was any guide. Not quite as active.
There's no such thing as "too connected." There *is* such a thing as "too willing to answer your connections."
The question of making yourself available is not a question of connectivity. Just because you carry a Treo or Sidekick or PocketPC Phone or whatever doesn't mean you are forced to answer emails now, today, tomorrow, or ever. Or that you have to pick up the phone. Or that you don't disable the ringer completely. Or that you even leave it on.
However... you can still be uber-connected, for your own benefit. Make calls when *you* want. Check your email on your schedule, not tied to when you're near a PC. Surf the web while you wait for your car at the shop. Connectivity is great, but making yourself immediately available to every inquiry is not, and there's a critical difference most people seem to be incapable of seeing.
Not only have they been saying they would develop it since 2002, they already have developed it, and T-mobile has refused to put it on their network for quite some time. It is deployed on several other hiptop providers' networks (Cable & Wireless and Edge, if I remember right) and works just fine.
Like another poster said-- this has been promised and re-promised and re-re-promised so many times and then held back so many different ways that I wouldn't count on sync until users are confirming it works for them. If you're considering one of these, and you need sync-- wait for it to be confirmed. Don't be burned like the rest of us were with promises of "sync to come" with no delivery to T-Mobile users for over two years, even when the software was finished, working, and deployed elsewhere.
The best way to get a good gamepad for the PC is just to use one of your console gamepads. You just need a USB adapter. I have a couple that support N64 and PSX gamepads, and use those, but they make them for just about everything. Here's a link to a place selling a hojillion varieties, including a 4-controller xbox-to-usb adapter for $20. You just plug it in and windows detects it and uses it like any other multi-axis analog gamepad. You'll need to do some button mapping, but that's it. $20, and use your existing favorite gamepads-- and you're off to the races!
Slashdot Mirror
Honestly, how long can it possibly take to sort 300 shows? It's just three hundred! I have a crappy java program here at work that sorts lists of thousands faster than you can say "I hate my tivo."
I would expect 300 shows to take longer than 20, but I'd expect 20 to take about 1/10 of a second. What can possibly be taking so long for the tivo to do?
Another question-- why are some people's boxes worse than others? Mine is slow, yours is not as slow, another poster says he's got no issue at all. A good friend of mine can't reorder his passes at all, except overnight, because it literally takes several hours to do. Why the variance?
I don't want to think about how slow it would be back with the default DirecTV guide. It may be faster, but it's still slower than molasses.
Interesting... I wonder what the variable is? I have a number of friends mad at me for selling them on the DirecTivo based on my Series 1 standalone tivo experience-- they're all locked into year-long contracts and having the same troubles I am, some worse. There's clearly a set of conditions that need to be met if you're not having the problems-- now we just need to figure out why yours works, and mine and my friends' don't.
I'll see if anybody's got aftermarket RAM for the units-- but I think it's soldered on the motherboard. And I'll go back to my original question-- how much RAM does it take to sort a list of 30 entries? I think my C64 could handle it nearly instantaneously. What is the tivo doing that takes more than an eyeblink?
Honestly, it sounds like a job for a first-year CS student. DirecTV has really bungled the Tivo software somehow. How long can it possibly take to reorganize a list of 30 entries? Or to read the guide database and print out 10 entries at a time? And if the backend stuff takes so long, why doesn't it get done in the background while the user goes back to watching TV? Even if there is a good reason that reordering your season passes takes 10 minutes, the UI should return control to the user and chug away in the background.
There's a 200MHz CPU in there, if I remember right. Why is it sooooo slow?
He needs to get into the DirecTV DVR code and figure out why it takes 30 seconds to display the guide, a minute to open your "Now Playing" list of shows, and 5+ minutes to sort a 30-entry list of season passes.
A huge fraction of Tivo's subscriber base is through the DirecTV tivos-- and despite my great experience with the standalone unit I had, the DirecTV box is so much slower despite 4x the processor speed that I can't even imagine what sort of horrible code is in there. Optimize the UI, *then* add features. DirecTV may singlehandedly turn millions of people away from tivo after they sign up and have a truly subpar experience with it.
The more primaries, the merrier. The gamut gets bigger the more you stick in there. A nice six-primary standard for displays would make me very happy... but so would a pocket-sized two-CCD digicam that could take 3D pictures. We can all dream.
As to Velvia, I understand it has a wider gamut than other film, but I haven't the faintest idea why. Perhaps the chemicals (and thus the color response) they've chosen correspond more closely to the response of our eyes? I'd be interested to know, if you've come across any information on what exactly makes Velvia so good.
My god I loved that game, and the spectacular sequel. Fighting ninjas in a trailer park while a tornado carries your trailer into the sky? Exploding robot kitten anti-personnel mines? Schoolgirl-aged assasins talking about going shopping before they try to chop your head off? A getaway sequence on a tricycle shooting evil mimes? Man-cubes as furniture? It goes on and on. I have never laughed so hard in my life. I cannot recommend NOLF and NOLF 2 enough.
You're certainly right about it being non-obvious. It took me a while to wrap my head around it, too.
The problem isn't that they've chosen the wrong primaries, although you're right that a different set could cover better than the ones we're using. The problem is that the three types of color sensors in the eye are not perfect single-frequency sensors-- they have wide, irregular frequency responses that overlap significantly. There are some spots where all three receptors would fire from a single-frequency beam of light, notably around 450-475, looking at the chart. How can you represent a different blue with a blue in that part of the spectrum if it fires the red (and/or green) receptors? It will inevitably look purple. The "negative" amount of red is just a mathematical artifact, a way to say "to represent this blue, you'd need a negative amount of red to cancel the phantom red caused by the blue light"-- and you simply can't do that in real life. As a photographer, you may have noticed effects like this occasionally in photographs. I have a shot of a beautiful deep blue flower I took in Arizona that is a crappy purple color in every shot I took. The film or CCD (both are RGB) you use to capture the image will produce the occasional noticable color distortion.
It's because of this overlap that we can't pick just three primaries that cover all the colors we can see. If the peaks were non-overlapping, it would work much, much better. So we need more complex models with nonlinear color mapping functions like CIE XYZ or CIE L*a*b to fully exploit the potential of the overlapping wide-response receptors in our eyes.
It is difficult to show with nothing but an RGB computer monitor, since it can't display the colors it can't display, but you might be able to find a Pantone book somewhere and look up some colors that are out of gamut for RGB displays, scan them or take a picture, and then look at the difference on your screen. You've probably noticed a similar effect when you try to print color images, as well, since color printing is CMYK, and not all RGB colors are in the gamut for CMYK.
D'oh! Well, scratch that part from what I said. I was unaware that CIE had used such a limited dataset in establishing the gamut. In my opinion, this is the sort of thing the HDTV standards board should have done-- a wide study to establish an accurate colorspace and delivery standard, so that we could have a nice, full-gamut (or as full as practical) display standard to go with our shiny new high-resolution televisions.
It's not the discrete gaps that are the problem! RGB does not represent all of the visible colors, even theoretically. Assuming a perfectly smooth RGB model with infinite intensity and perfect black, and infinitely precise levels of R, G, and B, there is a huge chunk (around 45%, if I remember right) of the visible gamut that is totally unreproducible. CMY covers some areas that RGB doesn't, and vice versa. Neither is the whole gamut. There are more complex models that do, like CIE L*a*b.
Not quite. The "step" issue is separate from the fact that the RGB gamut does not cover the visible gamut all the way. There are colors we can see that no amount of twiddling can *ever* get an RGB monitor to reproduce.
You are right that a digital RGB representation is discrete, not smooth, but there are colors "outside the grid," too. Pure yellow, for example.
Here's a nice link, again: clicky click
As others have pointed out, the RGB colorspace can't represent all of the colors.
RGB, CMY, CMYK, etc... *cannot* represent the entire visible color gamut. YIQ (the one used by NTSC TV), YUV (PAL TV), and YCrCb represent a smaller gamut than RGB, to be sure, but neither represent the whole thing.
For that, you need a more complex model like CIELAB.
Here's some links:
A whole lot of information.
Samsung stating that their shiny DTV sets can't match the visible gamut.
A graph of visible, RGB, Pantone, and CMYK gamuts
Quake 3 had used a physics model that did the calculations per-frame on the client side. To save CPU, this stuff isn't terribly precise, so there's some rounding error. Because of that, there were some "ideal framerates" where the rounding errors added up in your favor, resulting in the ability jump farther or run faster-- which is why the true nutcases built systems that could run at a framerate never below 125 (or whatever they believed/calculated the max to be), and then *locked* the framerate there.
p s. html
http://ucguides.savagehelp.com/Quake3/FAQFPSJum
has more info, along with a chart of where the best spots to lock your framerate are.
I'm not 100% sure, but I vaguely remember there being talk of a 60fps framerate lock in Doom III to prevent this sort of cheating.
People in every game work to get high framerates for other reasons, though-- really high *average* framerate means fairly high *minimum* framerate. And of course, minimums come right when you need it to be smooth-- like when 16 guys in jeeps, tanks, and airplanes all come down on you at once in BF1942.
It may seem absurd, but there are legitimate reasons why 3D cards need a higher framerate to represent the same smooth motion of 24fps movies.
I've explained this before, and I'll do it again. Television and video have motion blur-- the effect of the capture device essentially "averaging" the motion that occurred across the duration of capturing that frame.
Video cards generate a crisp, instantaneous frame that represents only the precise instant the frame was rendered, not the whole time the "shutter" was open.
At a *bare minimum* producing motion that looks as smooth as blurred 24fps requires double that. (You have to have two frames for your eye to blur between) To do it as well as a film camera requires even more, since their motion blur is effectively an infinite number of samples averaged together over the duration the shutter was open. I'd guess you could get a reasonable approximation at 3x the framerate.
TV and Movies are also filmed with the 24fps limitation in mind-- good cinematographers are well aware of the limits and know how to avoid situations that would result in jerky movement.
Tivos are a mixed bag. Series 2 units are fast and responsive compared to the Series 1 units, but DirecTV tivos (which are basically Series 2 units) are so slow that it can take 30 seconds to re-order your season passes, 5 seconds for the "now playing" list to pop up, etc...
Why you can't display or re-order a list of 30 text entries in less than a second with a 200MHz CPU baffles me. Sure, it won't play Doom III, but it really ought to respond more quickly to UI interactions. I'm not sure why the DirecTV Tivos are so awful-- it's the same hardware as the Series 2 Standalones.
Nicely done. I salute your metaphor-salvaging effort! But as the further nitpicking efforts by other posters point out, it probably ought to read something like:
:)
"Selling Cooling Machinery to the Indiginous Peoples Who Populate Areas Where It Is Always Colder Than The Lowest Temperature Obtainable Using The Machinery."
Which isn't as cool. And if it's really 54 where the Inuit are, and you can actually sell them Igloos, they might be interested in your special frozen-while-it's-above-freezing Igloo, which would make your original metaphor valid without any fudging.
I should just stop nitpicking. Everybody knew what you meant.
I believe you meant "selling iceboxes to eskimos." They might actually have a use for an igloo, as long as you can show that yours are a superior value over the homebuilt variety.
I should have posted longer, but was trying to keep it short. A plain old dremel worked almost as quickly, but needed two discs. We played around with it for quite a while after the first cut-- there's very little those locks protect against.
I don't think you'd have an issue carrying a battery-powered dremel and a dozen cutting discs to wherever you wanted to cut the lock.
Bolt cutters couldn't do it, to be sure. Abrasive cutting discs, no problemo.
Those "super-tough" Kryptonite U-locks take all of 15 seconds to cut through. I had a friend lock her wheel to the frame on her bike, and lose the key in a move. We were all set to resort to drastic measures with a blowtorch or liquid nitrogen, but all it took was compressed-air cutting tool (like a dremel, but powered like an impact wrench) with a proper cutting disc. Radiac, if I remember right. Went through it like butter. I've kinda lost all my faith in security products since then-- the first thing we tried went through it before you could say "hey, where's my bike?"
The shower of sparks might be a giveaway in a public place, though, so at least you've got that going for you.
Holy crap, somebody beat us to it! Ah well... I suppose the lesson is "no matter how dorky you you are, there's always someone dorkier."
Kudos to the folks setting this up-- it may be pointless, awkward to play, and a pain to set up, but that hasn't stopped me from wasting thousands of hours on similar futile excercises in geekery in the past, either.
Dance Dance Dragon's Lair Revolution
You're right-- it's the same damn game. "Memorize sequence of arbitrary steps/directions and enter them on the pad/joystick at exactly the right moment.
You could probably hook the dance pad right up to an old DL machine, and play the game that way. If there's a PC version of DL, I'll bet you could use a PC DDR pad with it out-of-the-box. Of course, the steps are farther apart, if my old amiga version was any guide. Not quite as active.
There's no such thing as "too connected." There *is* such a thing as "too willing to answer your connections."
The question of making yourself available is not a question of connectivity. Just because you carry a Treo or Sidekick or PocketPC Phone or whatever doesn't mean you are forced to answer emails now, today, tomorrow, or ever. Or that you have to pick up the phone. Or that you don't disable the ringer completely. Or that you even leave it on.
However... you can still be uber-connected, for your own benefit. Make calls when *you* want. Check your email on your schedule, not tied to when you're near a PC. Surf the web while you wait for your car at the shop. Connectivity is great, but making yourself immediately available to every inquiry is not, and there's a critical difference most people seem to be incapable of seeing.
Not only have they been saying they would develop it since 2002, they already have developed it, and T-mobile has refused to put it on their network for quite some time. It is deployed on several other hiptop providers' networks (Cable & Wireless and Edge, if I remember right) and works just fine.
Like another poster said-- this has been promised and re-promised and re-re-promised so many times and then held back so many different ways that I wouldn't count on sync until users are confirming it works for them. If you're considering one of these, and you need sync-- wait for it to be confirmed. Don't be burned like the rest of us were with promises of "sync to come" with no delivery to T-Mobile users for over two years, even when the software was finished, working, and deployed elsewhere.
The best way to get a good gamepad for the PC is just to use one of your console gamepads. You just need a USB adapter. I have a couple that support N64 and PSX gamepads, and use those, but they make them for just about everything. Here's a link to a place selling a hojillion varieties, including a 4-controller xbox-to-usb adapter for $20. You just plug it in and windows detects it and uses it like any other multi-axis analog gamepad. You'll need to do some button mapping, but that's it. $20, and use your existing favorite gamepads-- and you're off to the races!