>On a less practical note, it'd be quite interesting > if John Carmack would release the flight control >software he has developed for Armadillo's prototypes. > >While its highly proprietary code, it would give a >sense as to what needs to be tracked during the duration >of rocket flight and might interest others in >aerospace engineering.
I think in general we already know how to build flight controls for boosters. I'd be kind of interested, anyway. I've seen some pretty strange assertions on these topic from various shoestring-budget CATs designers. And, it's not directly a software problem - it's a lot more than just writing code.
>I'm entirely not convinced that was from a spy >satellite (to read 1-inch high lettering, the >targetting and stability problems alone would be >quite difficult to solve for such high resolution; >you'd have blurring (from spacecraft issues and >the person holding the book), mis-targetting, >etc.). Given that: > >All of the electronics have to be radiation >hardened. This usually puts back the technology by >a few years to even a decade compared with what >one could afford without the rad-hardening.
I don't know why this keeps coming up. In applications like this, computing power *is not* and *has not* been a limiting factor on spacecraft performance. Period. "Faster computers" have provided nearly no improvements in performance in applications like this. In fact, if you are really serious about high-bandwidth control systems you are still better off with *analog* and the requisite technology for that has existed for 50 years with negligible improvements. In fact, most if not all of the sensors (like earth sensors, star trackers, and any variety of gyroscopes) still use analog at the lowest level.
If anything, the advent of "better computers" and "better computer languages/programming practices" have probably *set the industry back* in terms of performance, and certainly set it back in the area of productivity. OO programming is probably great for some applications, but a control system implementation is essentially a procedural task. I've been in the business long enough to see the switch from analog/logic matrix hybrids, to procedural (done in FORTRAN, JOVIAL, and assembly) to OO. Some of the most efficient, clearly written, and maintainable code I've seen was implmented *using only IF statements and gotos*. Yes, you CAN write spaghetti code with FORTRAN, etc, and you CAN write clear and straightforward procedural code with C++. I've seen some absolutely incredible examples of both. But, if nothing else, in the good old days, you couldn't use the sort of stuff that you see in OO programming, because your GET and SET functions alone would suck up the entire memory and/or CPU. All that "better computers" have allowed is massive bloat, and associated explosion of questionably-applied OO programming. For this application the desired level of abstraction is the *bit*. But I feel another rant coming on...
More computing power and digital flight control systems provide much more flexibility and more easily-implemented features - but they DO NOT necessarily have anything to do with improving pointing performance.
In any case, the limiting factor in getting high-resolution has absolutely nothing to do with rad-hard technology dragging down performance. Sufficient controls performance can be acheived without computers at all, and was possible and achieved in the 60's
Structural exitation (jitter, bending) IS a limiting factor on performance, and most of the items that need to point some device accurately are designed with this in mind. But it's always a tradeoff between rigidity/damping and weight.
In any case, the ultimate limiting factor on the resolution is the size of the objective (almost always a mirror), and there's only so much glass you can launch to Mars with a relatively inexpensive rocket. You want to double the resolution, come up with 10x the money, and I'm sure we can figure out a way to get it.
> I would not be surprised if internet radio over Wi-Max ends up > seriously hurting XM and Sirius. Right now they have a > stranglehold on the market due to the cost of entry (satelite > systems). However Wi-Max enabled car "radios" open up a > whole new realm of possibilities, for both competing subscription > channels and free alternatives
I would be very surprised if WiMax or any other ground-based streams will significantly effect satellite radio. Unless they put WiMax on a satellite, there will be no connectivity over VAST reaches of the US. There are large areas of this country in which you can't even get AM radio - and I mean *no stations*, not just stations you want to listen to. That's after close to a century of development.
I doubt that streaming digital data will be any more available. That's by far the biggest appeal of Sirius - you can literally go anywhere in the country and get exactly the same service.
> I believe this is important. To learn more about long flights is indeed useful. > Also how long you can fly before the fuel weight works against you, how fast > you can fly in order to get the longest distance etc.
These were all burning questions - in 1935. Charles Lindberg did exactly this sort of analysis in 1927 and the basic concepts are completely unchanged since then. The only reason this couldn't have been done in the 30's is that the engines weren't reliable enough.
No offense to anyone, but to suggest that Fawcett is discovering some new principles in aeronautics by doing this flight is absurd. This type of information has been perfectly well understood for at least half a century. It's a neat stunt, but that's all it is.
>Anyone else worried about militaries getting even more involved in space?
Wow. Clearly, you don't understand the space business - at all. Military and other related agencies are *the primary users* of space resources, and have been from the beginning. NASA, commercial, and other civilian users are very minor players in aerospace. Only during the mid-60's with Apollo was it even comparable in any terms.
Point being, military/related users ARE the space business. Cadets tossing together a satellite is "in the noise" of the existing military uses.
>In a murder case, should the judge always believe the suspect > when he claims "not guilty"?
" Believe" doesn't enter into it. The judge is *required* to treat the suspect as innocent until proven otherwise. This ain't France.
And, in case it has escaped the laser-like focus of the local self-declared legal experts, the "evidence" that such a a wiretapping situation exists in the first place *was illegally obtained*. A felony was committed. We don't know who committed it, but it is unequivocally a federal felony to intentionally disclose classified information. But no one here seems to be too concerned over that crime.
>Furthermore, they left out the fact that on the Dual Shock 2, > the four buttons (triangle, square, circle, X) are >velocity/pressure-sensitive. In other words, games can (if they are programmed to > do so) tell how hard you pushed the buttons. This feature isn't used > in too many games,
The arrow pad is also pressure-sensitive.
All these are definitely used in GT3/GT4. In fact, using the pressure sensitivity on the throttle is vital in some parts of the game. If you don't use it you will burn the tires off in half a lap and/or crash into every available stationary object. Of course, it's tricky to use since the range of pressures is super-tiny from barely on to full blast. Steering with the arrow pad is borderline hopeless in GT4 and just functional in GT 3.
Insulation on the inside was done on the S-IVb. The only reason you couldn't do it for the shuttle is that you would have to completely redesign the tank.
> Surely you would be able to hear the limitations >of that format on that kind of equipment.
You can hear it on your iPod, for that matter. Doesn't take a "golden ear" or anything special. As you get better equipment, all the other issues start becoming audible, too. For most pop-type recordings, 128 to 192 is about where the break point between encoding method and the other recording limitations lies.
>but can you imagine if Clearwire instituted this policy? Or XM radio?
I can imagine it. It would mean that XM would, effectively, default on their user contracts. And within a month, be on the inexorable slide to bankruptcy when all the backers (who have been keeping it alive so far) back out. That's not to disrespect indepent labels and artists, but it's still far from clear that satellite radio is financially viable even with everything. The RIAA is a virtual monopoly, and if satellite radio couldn't play pop pablum for the masses they would certainly not be able to make a go of it. Sirius would have a chance only because of the Howard Stern show, but, I would point out, they haven't made any money yet, either.
I haven't listened to XM much, but I can confirm the general problem. Sirius has quite low sound quality, just barely on the edge of acceptable. I did a pretty extensive test of Sirius via FM modulator in the car, via direct input in the car, direct input at home, and the Internet streaming audio. While there were obvious differences in the different methods, none of them approach even iTunes quality, much less CD. There's a sort of hollow metallic reverb in *all* delivery methods. After going round and round with Sirius, I came to the conclusion that it's inherent in thier system. And it is quite grating on some of the lower-bandwidth channels. Talk channels make AM radio sound like nirvana.
But they were never selling quality, they are selling convenience. And it is amazingly convenient. Unless, as I have, you get receivers that don't work for more than about 1/2 an hour, in which case it's a pain in the ass.
I still have the service, and when everything works it is just worth the $13 a month. I actually make use of the Internet stream more than I use the satelllite downlink.
You point is pretty much right on the nose- recording satellite radio is not going to make most people stop buying CDs because the quality is too low. BUT, the RIAA really only cares about trying to suck money out of the system in every possible way. Even if the basis for concern is obviously wrong, that won't stop them from trying.
> Most people couldn't fathom of filling up 60GB with music, > particularly paid music, so either they're clueless about > purchasing decisions, or they're buying the large Ipods > for something else, namely prestige.
I call shenanigans! If you encode at 128 Kbps, then sure, you can maybe fit "500 songs" on your Nano. But it you encode it as lossless from CDs, 60 gig can be sucked up in a surprisingly short time. And, even with the regular earbuds, almost anyone who really paid attention could tell the difference in the sound quality. If you listen critically with decent headphones, the difference is night and day.
The Russians got even further along this kind of thinking. Look into the "E-4 project" (I won't post my favorite link because while it might help my karma *here* it certainly won't help with my karma with the server owner to/. him!)
It was actually not that bad an idea, before the advent of reliable complex telemetry systems, it was a fairly reasonable way to prove you had done it.
Sucks is a strong word, but hi-fi it ain't. I went back and forth with Sirius over the sound quality, and they were pretty responsive and helpful, but they couldn't fix it. Lots of assuming that "my equipment was bad" - I was discussing my home system and mentioned the various high-end (real high-end, not 'audiofile' "good-sounding speaker wire and shakti stones" crap) parts, like my semi-custom AVA amplifier, and they mentioned that I might "upgrade" to Pioneer or Yamaha, for example. But the bottom line was the sound issues are a function of the encoding. It's not very good. The kicker was when I hooked up to the streaming audio over the internet and it was absolutely no different than the broadcast sound.
It's *very* obvious in the DJ speaking voices, but it happens in the music similarly. The worst is a sort of a "hollow" reverb effect.
By the way, the best feature IS the streaming audio, which is free if you subscribe to the broadcast service (or, is included, if you prefer that perspective).
I listened to mine on broadcast for 6 straight days on a car trip, and I had a lot of opportunity to compare it to FM stations along I-80. The best Sirius channels are nowhere near as good in terms of audio quality as a good FM station, and the talk channels are worse than AM. I tried various encoding schemes from CD to compare, and somewhere in the range of a 96 kbps MP3 was pretty comparable to the very best Sirius channels. In other words, just barely good enough for most people, and not a whole lot worse from what a lot of people tolerate on their iPods (128 kbps is what I think you get from the ITMS - whatever it is it's on the edge of tolerable quality-wise). Which I guess is what they were shooting for.
couldn't even find a bit rate low enough to replicate the worst of the talk channels.
I think it's *probably* worth the money, but if you are expecting CD quality sound you will be sorely disapppointed.
I think there are an awful lot of excessively harsh comments here. These guys are trying to do a very difficult mission with virtually no applicable experience. Sure, there's a lot of information that, say, NASA (or more accurately, the contractors who work for NASA/DOD/other customers) has, but you simply can't effectively transfer the experience needed for this type of engineering.
By the way, there are some very good reasons* that US space projects take so long and cost so much. It's because there are enough experienced engineers around to know where to worry and what to worry about. That didn't build up in a few tries. Some contractors have literally hundreds, bordering on a thousand, successful space missions under their belts.
All the documents and "design processes" in the world cannot make up for having a few guys around with the necessary background to take a look at the design or implementation, and just see where the obvious problem areas might me.
It's a tough loss, but it's just the cost traversing of the learning curve.
Brett
* Of course, there are some "bad reasons" for overruns and schedule slips, too - usually, overly-optimistic schedules and budgets! In fact, a lot of the time there is no one in the entire acquisition system that has a stake in getting the *real* cost or *real* schedule.
Re:More Bad News: No Vorbis
on
VW Goes USB
·
· Score: 1
>``The bad news? I just got my Touran... sans >USB.'' > >That, and I bet it doesn't support Ogg Vorbis. I >understand this is because of lack of consumer >demand and visibility, but it still hurts me that >support for an open, royaltee-free and superior >format is so utterly lacking.
Not to be a wise guy, but supporting a format that quite literally *no one* outside a small group of hard-core OSS programmers have ever heard of would be completely absurd.
In any case, the correct answer to the "device compatibilty" issue (as mentioned in other responses) is to instead provide analog line in. It's not as clean as "digital" but if properly implemented the signal degradation will be completely beyond anyone's ability to hear, even if it were sitting in your living room. If you hear artifacts from using analog inputs vice digital, the analog is not implemented correctly.
And need I point out that we are talking about listening to music IN A CAR - i.e. a glass/metal box connected to an internal combusion engine and whzzing down the highway at 80 mph. The fidelity requirements for an audio system in those conditions are not daunting.
Slashdot Mirror
>On a less practical note, it'd be quite interesting
> if John Carmack would release the flight control
>software he has developed for Armadillo's prototypes.
>
>While its highly proprietary code, it would give a
>sense as to what needs to be tracked during the duration
>of rocket flight and might interest others in
>aerospace engineering.
I think in general we already know how to build flight controls for boosters. I'd be kind of interested, anyway. I've seen some pretty strange assertions on these topic from various shoestring-budget CATs designers. And, it's not directly a software problem - it's a lot more than just writing code.
Brett
I think you are referring to this:
http://genesismission.jpl.nasa.gov/
The probe crashed because an accelerometer was upside down, but the data was mostly unaffected. It was September 2004, so, not that long ago.
Bre
Bingo, couldn't say it better myself.
Brett
>Isn't all this stuff written in assembly anyways? Ewww digging through
> xor's and and's and sll's fun stuff!
Almost certainly not. Probably C++, but perhaps Ada.
Brett
>I'm entirely not convinced that was from a spy
>satellite (to read 1-inch high lettering, the
>targetting and stability problems alone would be
>quite difficult to solve for such high resolution;
>you'd have blurring (from spacecraft issues and
>the person holding the book), mis-targetting,
>etc.). Given that:
>
>All of the electronics have to be radiation
>hardened. This usually puts back the technology by
>a few years to even a decade compared with what
>one could afford without the rad-hardening.
I don't know why this keeps coming up. In applications like this, computing power *is not* and *has not* been a limiting factor on spacecraft performance. Period. "Faster computers" have provided nearly no improvements in performance in applications like this. In fact, if you are really serious about high-bandwidth control systems you are still better off with *analog* and the requisite technology for that has existed for 50 years with negligible improvements. In fact, most if not all of the sensors (like earth sensors, star trackers, and any variety of gyroscopes) still use analog at the lowest level.
If anything, the advent of "better computers" and "better computer languages/programming practices" have probably *set the industry back* in terms of performance, and certainly set it back in the area of productivity. OO programming is probably great for some applications, but a control system implementation is essentially a procedural task. I've been in the business long enough to see the switch from analog/logic matrix hybrids, to procedural (done in FORTRAN, JOVIAL, and assembly) to OO. Some of the most efficient, clearly written, and maintainable code I've seen was implmented *using only IF statements and gotos*. Yes, you CAN write spaghetti code with FORTRAN, etc, and you CAN write clear and straightforward procedural code with C++. I've seen some absolutely incredible examples of both. But, if nothing else, in the good old days, you couldn't use the sort of stuff that you see in OO programming, because your GET and SET functions alone would suck up the entire memory and/or CPU. All that "better computers" have allowed is massive bloat, and associated explosion of questionably-applied OO programming. For this application the desired level of abstraction is the *bit*. But I feel another rant coming on...
More computing power and digital flight control systems provide much more flexibility and more easily-implemented features - but they DO NOT necessarily have anything to do with improving pointing performance.
In any case, the limiting factor in getting high-resolution has absolutely nothing to do with rad-hard technology dragging down performance. Sufficient controls performance can be acheived without computers at all, and was possible and achieved in the 60's
Structural exitation (jitter, bending) IS a limiting factor on performance, and most of the items that need to point some device accurately are designed with this in mind. But it's always a tradeoff between rigidity/damping and weight.
In any case, the ultimate limiting factor on the resolution is the size of the objective (almost always a mirror), and there's only so much glass you can launch to Mars with a relatively inexpensive rocket. You want to double the resolution, come up with 10x the money, and I'm sure we can figure out a way to get it.
Brett
Boy, I really hope somebody got fired for that blunder!
http://www.snpp.com/episodes/4F12.html
Brett
> I would not be surprised if internet radio over Wi-Max ends up
> seriously hurting XM and Sirius. Right now they have a
> stranglehold on the market due to the cost of entry (satelite
> systems). However Wi-Max enabled car "radios" open up a
> whole new realm of possibilities, for both competing subscription
> channels and free alternatives
I would be very surprised if WiMax or any other ground-based streams will significantly effect satellite radio. Unless they put WiMax on a satellite, there will be no connectivity over VAST reaches of the US. There are large areas of this country in which you can't even get AM radio - and I mean *no stations*, not just stations you want to listen to. That's after close to a century of development.
I doubt that streaming digital data will be any more available. That's by far the biggest appeal of Sirius - you can literally go anywhere in the country and get exactly the same service.
Brett
Correct, in fact, CDs are about 1.4 Mbps. 16 bits at 44.1 Khz, on two channels.
Brett
> I believe this is important. To learn more about long flights is indeed useful.
> Also how long you can fly before the fuel weight works against you, how fast
> you can fly in order to get the longest distance etc.
These were all burning questions - in 1935. Charles Lindberg did exactly this sort of analysis in 1927 and the basic concepts are completely unchanged since then. The only reason this couldn't have been done in the 30's is that the engines weren't reliable enough.
No offense to anyone, but to suggest that Fawcett is discovering some new principles in aeronautics by doing this flight is absurd. This type of information has been perfectly well understood for at least half a century. It's a neat stunt, but that's all it is.
Brett
"I gave him some special ointment and he hurt so bad that he had to apply it in the car with his sock"
Brett
>Anyone else worried about militaries getting even more involved in space?
Wow. Clearly, you don't understand the space business - at all. Military and other related agencies are *the primary users* of space resources, and have been from the beginning. NASA, commercial, and other civilian users are very minor players in aerospace. Only during the mid-60's with Apollo was it even comparable in any terms.
Point being, military/related users ARE the space business. Cadets tossing together a satellite is "in the noise" of the existing military uses.
Brett
>In a murder case, should the judge always believe the suspect
> when he claims "not guilty"?
" Believe" doesn't enter into it. The judge is *required* to treat the suspect as innocent until proven otherwise. This ain't France.
And, in case it has escaped the laser-like focus of the local self-declared legal experts, the "evidence" that such a a wiretapping situation exists in the first place *was illegally obtained*. A felony was committed. We don't know who committed it, but it is unequivocally a federal felony to intentionally disclose classified information. But no one here seems to be too concerned over that crime.
Brett
So you assume that the wiretaps were/are illegal. Although that
is far from an established fact. I mean, the NYT said so, it must
be true...
Brett
>Furthermore, they left out the fact that on the Dual Shock 2,
> the four buttons (triangle, square, circle, X) are
>velocity/pressure-sensitive. In other words, games can (if they are programmed to
> do so) tell how hard you pushed the buttons. This feature isn't used
> in too many games,
The arrow pad is also pressure-sensitive.
All these are definitely used in GT3/GT4. In fact, using the pressure sensitivity on the throttle is vital in some parts of the game. If you don't use it you will burn the tires off in half a lap and/or crash into every available stationary object. Of course, it's tricky to use since the range of pressures is super-tiny from barely on to full blast. Steering with the arrow pad is borderline hopeless in GT4 and just functional in GT 3.
Brett
Insulation on the inside was done on the S-IVb. The only reason you couldn't do it for the shuttle is that you would have to completely redesign the tank.
Brett
> Surely you would be able to hear the limitations
>of that format on that kind of equipment.
You can hear it on your iPod, for that matter. Doesn't take a "golden ear" or anything special. As you get better equipment, all the other issues start becoming audible, too. For most pop-type recordings, 128 to 192 is about where the break point between encoding method and the other recording limitations lies.
Brett
>but can you imagine if Clearwire instituted this policy? Or XM radio?
I can imagine it. It would mean that XM would, effectively, default on their user contracts. And within a month, be on the inexorable slide to bankruptcy when all the backers (who have been keeping it alive so far) back out. That's not to disrespect indepent labels and artists, but it's still far from clear that satellite radio is financially viable even with everything. The RIAA is a virtual monopoly, and if satellite radio couldn't play pop pablum for the masses they would certainly not be able to make a go of it. Sirius would have a chance only because of the Howard Stern show, but, I would point out, they haven't made any money yet, either.
Brett
I haven't listened to XM much, but I can confirm the general problem. Sirius has quite low sound quality, just barely on the edge of acceptable. I did a pretty extensive test of Sirius via FM modulator in the car, via direct input in the car, direct input at home, and the Internet streaming audio. While there were obvious differences in the different methods, none of them approach even iTunes quality, much less CD. There's a sort of hollow metallic reverb in *all* delivery methods. After going round and round with Sirius, I came to the conclusion that it's inherent in thier system. And it is quite grating on some of the lower-bandwidth channels. Talk channels make AM radio sound like nirvana.
But they were never selling quality, they are selling convenience. And it is amazingly convenient. Unless, as I have, you get receivers that don't work for more than about 1/2 an hour, in which case it's a pain in the ass.
I still have the service, and when everything works it is just worth the $13 a month. I actually make use of the Internet stream more than I use the satelllite downlink.
You point is pretty much right on the nose- recording satellite radio is not going to make most people stop buying CDs because the quality is too low. BUT, the RIAA really only cares about trying to suck money out of the system in every possible way. Even if the basis for concern is obviously wrong, that won't stop them from trying.
Brett
> Most people couldn't fathom of filling up 60GB with music,
> particularly paid music, so either they're clueless about
> purchasing decisions, or they're buying the large Ipods
> for something else, namely prestige.
I call shenanigans! If you encode at 128 Kbps, then sure, you can maybe fit "500 songs" on your Nano. But it you encode it as lossless from CDs, 60 gig can be sucked up in a surprisingly short time. And, even with the regular earbuds, almost anyone who really paid attention could tell the difference in the sound quality. If you listen critically with decent headphones, the difference is night and day.
Brett
Could you please explain what sort of ecology exists on the moon to be destroyed?
Brett
The Russians got even further along this kind of thinking. Look into the "E-4 project" (I won't post my favorite link because while it might help my karma *here* it certainly won't help with my karma with the server owner to /. him!)
It was actually not that bad an idea, before the advent of reliable complex telemetry systems, it was a fairly reasonable way to prove you had done it.
Brett
"Everybody Loves Hypnotoad"
We can only hope...
Brett
Sucks is a strong word, but hi-fi it ain't. I went back and forth with Sirius over the sound quality, and they were pretty responsive and helpful, but they couldn't fix it. Lots of assuming that "my equipment was bad" - I was discussing my home system and mentioned the various high-end (real high-end, not 'audiofile' "good-sounding speaker wire and shakti stones" crap) parts, like my semi-custom AVA amplifier, and they mentioned that I might "upgrade" to Pioneer or Yamaha, for example. But the bottom line was the sound issues are a function of the encoding. It's not very good. The kicker was when I hooked up to the streaming audio over the internet and it was absolutely no different than the broadcast sound.
It's *very* obvious in the DJ speaking voices, but it happens in the music similarly. The worst is a sort of a "hollow" reverb effect.
By the way, the best feature IS the streaming audio, which is free if you subscribe to the broadcast service (or, is included, if you prefer that perspective).
I listened to mine on broadcast for 6 straight days on a car trip, and I had a lot of opportunity to compare it to FM stations along I-80. The best Sirius channels are nowhere near as good in terms of audio quality as a good FM station, and the talk channels are worse than AM. I tried various encoding schemes from CD to compare, and somewhere in the range of a 96 kbps MP3 was pretty comparable to the very best Sirius channels. In other words, just barely good enough for most people, and not a whole lot worse from what a lot of people tolerate on their iPods (128 kbps is what I think you get from the ITMS - whatever it is it's on the edge of tolerable quality-wise). Which I guess is what they were shooting for.
couldn't even find a bit rate low enough to replicate the worst of the talk channels.
I think it's *probably* worth the money, but if you are expecting CD quality sound you will be sorely disapppointed.
Brett
I think there are an awful lot of excessively harsh comments here. These guys are trying to do a very difficult mission with virtually no applicable experience. Sure, there's a lot of information that, say, NASA (or more accurately, the contractors who work for NASA/DOD/other customers) has, but you simply can't effectively transfer the experience needed for this type of engineering.
By the way, there are some very good reasons* that US space projects take so long and cost so much. It's because there are enough experienced engineers around to know where to worry and what to worry about. That didn't build up in a few tries. Some contractors have literally hundreds, bordering on a thousand, successful space missions under their belts.
All the documents and "design processes" in the world cannot make up for having a few guys around with the necessary background to take a look at the design or implementation, and just see where the obvious problem areas might me.
It's a tough loss, but it's just the cost traversing of the learning curve.
Brett
* Of course, there are some "bad reasons" for overruns and schedule slips, too - usually, overly-optimistic schedules and budgets! In fact, a lot of the time there is no one in the entire acquisition system that has a stake in getting the *real* cost or *real* schedule.
>``The bad news? I just got my Touran... sans
>USB.''
>
>That, and I bet it doesn't support Ogg Vorbis. I
>understand this is because of lack of consumer
>demand and visibility, but it still hurts me that
>support for an open, royaltee-free and superior
>format is so utterly lacking.
Not to be a wise guy, but supporting a format that quite literally *no one* outside a small group of hard-core OSS programmers have ever heard of would be completely absurd.
In any case, the correct answer to the "device compatibilty" issue (as mentioned in other responses) is to instead provide analog line in. It's not as clean as "digital" but if properly implemented the signal degradation will be completely beyond anyone's ability to hear, even if it were sitting in your living room. If you hear artifacts from using analog inputs vice digital, the analog is not implemented correctly.
And need I point out that we are talking about listening to music IN A CAR - i.e. a glass/metal box connected to an internal combusion engine and whzzing down the highway at 80 mph. The fidelity requirements for an audio system in those conditions are not daunting.
Brett