Yes. There was a lot of ham software written for the PDP-8, 10, and 11, and much of this was ported to the 8080 and 6800 in 1975/1976.
Never discount the value of dusty decks and Code Inertia.
Frankly, nobody in their right fleeping mind does honest realtime work on Linux; if anything, the various tiny BSD kernels would be the tool of choice. Most realtime code these days is written either for nuxi-like environments like VxWorks, or hand bummed for custom threaded operating systems.
I wouldn't exactly proclaim ham radio as a bearer of tech progress either. As I stated above, the innovations have been more economic than technical at times. But give credit where credit is due. Where you around for the Sheet Metal Days of microcomputing? How do you think the hardware community communicated and swapped fixes back then? Yes, you could wait for the bi-monthly flimsy newsletter that some poor hack put out on an irregular basis. Or you could instead interact with the weekly net on 20 meters, ask if anyone had seen a problem or had a fix, and have them describe it to you directly. The choice is pretty obvious.
That's probably because computing in ham radio predates the Free/Open Software movements by more than a decade. Most of the first microcomputer hobbyists were hams, because amateur radio shortwave nets allowed people to swap experiences and help each other out. And of course, many of the first applications written were for the ham shack, from RTTY terminals to logbooks to propagation forecasting to satellite orbital prediction.
As time wore on, programs written for the x80-CP/M environment were ported to the x86-DOS world because that's what was available, and most radio applications required what the PC gave you -- the whole machine. But, in all this time, there was never a heritage of code-sharing, since making a little money on the side allowed you to defray the costs of the hobby. That's one well-known characteristic of hams -- they're cheapskates by nature or pick it up as they go on -- and as a result many ham radio innovations are economic, not technical.
As more modern multitasking operating systems showed up, non-realtime apps have been ported to them, and some amazing semi-realtime DSP work has been done, such as the various PSK31 implementations. And many of these are at least of an Open philosophy -- almost all of the PSK31 implementations, for example, are based on a single core DLL produced under a "share and enjoy" license.
So indeed, amateur radio comes from the DOS world, and the Free Nuxis have a lot of catching up to do.
Like any peripheral, there are two basic approaches to interfacing with the beast:
1. Internal. This is the Winradio approach. The good news is that, as processors become faster and faster, they're able to absorb more of the electronics into software layers. The latest WinRadio is akin to all those Winmodems we've seen. However, the environment within a computer case isn't exactly the best place to put an RF circuit -- it's full of all kinds of strong fields and oddball harmonics.
2. External. A much nicer place to put your radio is in a nice RF tight box a few feet away from all that nonlinear harmonic crapola. And, after all, the output is relatively low bandwidth, so bring it into the system through an I/O port -- USB, 1394, heck even a serial port will do.
What you really want is an Icom PCR-1000, covers 100 kHz to 1.3 GHz (continuous if you shop in Akihabara), multimode. Hook it up to a serial port and an audio in jack, and you're all set to vacuum the ether.
Or, just check out the JavaRadio network of PCR-1000 equipped sites around the world...
...the last remaining religion that atheism can bash with impunity. Anything else would be antisemitic, or a hate crime. Even Islam sees more politically correct protection these days ("mustn't incite the people to war, you know...").
There are good, sound, moral reasons for opposing embryonic stem cell research, if you have any respect for life or the living that a belief in a Deity has nothing to do with. But, rather than address the arguments logically, Reeves has instead chosen the Ad Hominem. Swell.
While Intel and AMD have been locked in the performance race, Cyrix took their x86 line in a different direction: low power consumption and low heat dissipation. I just built a new box based on the VIA/Cyrix Eden ITX formfactor motherboard for less than $400, and it consumes less than 36w (no fan). And, best of all, it runs everything in the ham shack -- including some powerful DSP software (PSK31).
Yes, but let's face it: none of these shows holds a candle to the true King of Implied Violence Entertainment: Jerry Springer. Thing is, if you tune into Springer, there's an expectation of (at least thinly veiled) violence. As far as I'm able to tell, Phil Donohue hasn't hired bouncers...yet.
But then again, there's always been a certain thrill to heated arguments. Anybody remember Point / Counterpoint on 60 Minutes? [Okay, we *all* remember the SNL parody...] So this is nothing new.
The Earthquake Experience varies greatly depending on where you are at the time. If you're up in the hills, particularly the granite outcroppings in San Mateo county, you'll barely feel a thing. On the other hand, if you're down in the flats, especially east of Highway 101, even a little 3.0 will get your attention. I was out in the "bay alluvial mud flats" part of Palo Alto for the 1989 quake, had these three foot waves in my waterbed. Something to keep in mind if you're investing in an oh-my-stars-that's-expensive house out here.
Right now, Caltrain is running at Restricted Speed south of Redwood Junction, I'm listening to the dispatcher dish out track-n-time as the track and bridges get inspected...
You can only simulate what you understand, and more often than not, simulations tend to forget some minor detail that invalidates the result. This is why the tendency of certain pseudoscience groups to hide behind simulations is appalling, since minor details often determine the outcome in Real Life.
I sincerely doubt, for example, that the author of this study took into account motion of the "point sources". Next time you ride a train, look at how people tend to rock back and forth as the suspension dampens the shock of rail joints and other track discontinuities. It's not much, maybe five to ten centimeters, but that's enough to cancel out constructive interference when the wavelength is only about 16cm.
This is why experimental science and empirical engineering are so crucial; they either confirm our understanding of the problem, or it's back the drawing board....
"And you wonder why CDMA cell phones are much larger than their GSM counterparts".
Yes, I would wonder that, if there were any truth to it. My Motorola CDMA Vader is just as small as the GSM Vader. And I have an 802.11b card (Symbol's CF Wi-Fi) that's smaller than any GSM phone. Don't overgeneralize.
When it comes down to it, Spread Spectrum is nothing more than a spectrum allocation algorithm that depends on the orthogonality of the chipping sequences. This orthogonality is further degraded by Rayleigh fading (aka multipathing), which is why ACPR or diversity reception is essential to CDMA. The primary benefit, however, is that that as an allocation scheme, spread spectrum is time based, and can take advantage of the bursty nature of most communications, be they human-to-human or machine-machine. And that's why time domain approaches are more efficient.
Divvying up the spectrum is a concept that is as old as ham radio, and almost as outdated. The whole concept of spectrum allocation is being turned on its ear by new technology which now allows for coherence in the time domain instead of the traditional frequency domain approach. And yes, there has been quite a bit of work on spread spectrum done on amateur frequencies, and yes, this does raise the noise floor for traditional narrowband users.
The amateur service is intended to be more than merely a backup yacking system to cellphones. It's supposed to encourage experimentation with new systems. Sadly, the current FCC rules don't allow for much useful spread spectrum experimentation by hams, but the truth of the matter is that existing commercial systems could be ported to the amateur frequencies, and do the jobs you mention (emergency communications) far, far better than the 1950s FM technology you're clinging to.
Look at what happened to Packet Radio in the Amateur Service; we built it, they came, but then the FCC issued and enforced inane and archaic content restrictions on the use of packet, and the experimentation died out. The FCC is doing it to us again, all in the name of control.
...and a damn fine one, at that. Rather, it's that Richard has a lot of knowledge about how to use AltaVista, both by personal experimentation and by knowing people who worked on the project. His book was one of the first really great compilations of how to use a search engine to solve problems other than "what web page has info on this topic?". So, no, Richard doesn't "make his living because of AltaVista" by a long shot.
Cory Doctrow's article, BTW, is riddled with errors and false assumptions about the ranking algorithms used by both Google and AltaVista, which have evolved radically since their introduction. AltaVista, for example, ceased to be a purely keyword-based engine even before it hit the street. And, now that Google actually has a substantially sized audience, they're having to deal with a lot more spam attacks than ever before.
The real battle to come isn't going to be entirely technology based, however. It's going to be a duel between CMGI and Kleiner-Perkins to see who can figure out how to make a profitable (with good ROIC) search engine business before AltaVista and Google drain their coffers. In that aspect, Google is woefully behind.
If you're 26 years old, then you're too young to remember the Cold War. You know, the one we WON because nobody in their freaking mind would attack the US military? Well, guess what: eight years of military softening by the Philanderer-in-Chief left not only our military but our CIVILIAN population vulnerable to some psycho with a Weapon of Mass Destruction (and yes, a 767 at speed loaded with fuel qualifies as a WMD).
There are only two options for dealing with such people. One is to eliminate any civil liberties, and spot them as they're about to attack. The other is to kill them in place. I'll opt for the "Kill in Place" method, since I see no reason to screw up a perfectly good society.
As for the rest of the world, they do what's in their best interest without consulting the US. Get a clue.
No, the biggest problem with 802.11a is that it isn't spread spectrum. Yes, I know that some adherents like to think that FH is SS, but a quick look at a spectrum analyzer will show you how unspread the signal is, full of peaks and impulses. And then there's the 5 GHz question. Four times the path loss, and unwanted gain.
Face it, backward compatibility matters. And for the broader market, that means 2.4 GHz and something that will do "b" in a pinch. One can argue technical merits until you're blue in the face, but what really matters is the market, and if the history of POTS modem development is any guide, "g" is the future.
Three researchers at DEC's Palo Alto Labs decided to build a search engine to show that DEC did the Internet and Databases better than anyone else. When they finished, they showed it to the management, who didn't have a clue as to what to do with it. The problem was that DEC was trying very hard to sell itself to Compaq. They didn't want to invest capital in new ideas, since that would shink the cash-on-hand and make them appear worth less to Compaq. As a result, AltaVista was always cash poor and couldn't afford to buy hardware (from DEC!) to keep up.
The first spinout attempt was lead by a CEO who wanted AltaVista to be a software company, and was ended by the 1997 market plunge. Then Compaq bought DEC, and eventually sold most of AltaVista to CMGI. They then attempted to convert AltaVista into a portal, since the investment-driven strategy of the day called for eyeballs and land grab. A huge amount of cash was spent on the portal strategy (and very little on the search engine) until the day in April 2000 when eyeball valuations died.
The moral of the story is that market and capital strategy matter just as much as technology.
Many businesses are now failing, not because of bad technology or bad business models, but because of poor execution. Startups are particularly beholden to their early investors, and all of the VCs were heavily pushing "land grab" business practices throughout the bubble. The reason is obvious: Stocks were valued primarily on audience, and if you want the best quick return on your investment, you demand practices that build audience share, period. Early investors don't care about long term prospects for a business, they just want a quick cash-out to flip into something else. The 1997 changes in post IPO-lockout (from two years to six months) only magnified the short term focus. Having structured the business with that bias, the result is inevitable.
Businesses fail for all kinds of reasons, many of which are completely out of control of the management or techs. Don't dismiss an idea just because somebody screwed the pooch in the past.
Give the world's best search engine to someone who only knows how to sell consumer PCs, and who brags that he "doesn't know anything about the Internet".
Give said CEO a billion in cash to waste on bad acquisitions and an unneeded advertising campaign in order to boost the stock valuation.
Stand by as $3 billion in value is destroyed by lack of capital investment in the core business.
Dance around the fire at night chanting "Smart is Beautiful!"
Alpha was the great architecture that blew everything else away, but thanks to the guys in Houston, we're left with the One True Architecture. Bah!
"The 80xxx series of microprocessors is clear evidence that INTEL isn't doing in-house
drug testing." (Usenet, 1988)
"It would be different if Slashdot made any claims to political, social, or cultural neutrality, but the fact is that it doesn't. If you read this site, you should expect a liberal bent. If you want a rightwing take on the same issues, I'm sure Google will dig up plenty of suitable articles and discussions. "
Spoken like a true soldier of Political Correctness. "If you're in the University, you should expect a liberal slant, and if you don't like it, go someplace else!". Yeah, right. Slashdot proclaims itself as News for Nerds. Last time I checked, Engineers were for the most part considered Nerds, and more than three out of four Engineers consider themselves politically conservative. You'll find just the opposite numbers for Computer Scientists (and leaning but not quite as far left for most of the physical sciences). Slashdot, for all intents and purposes, is News for CS Types , and the resulting political bent comes as something of a shock for those who haven't figured out that Computer Scientists are not Engineers!
Actually, up until about 1930 or so, most innovation in the US *was* privately funded -- because the federal government lacked the money before the passage of the income tax as a method of confiscating wealth. Companies like RCA, Philips, US Steel, Westinghouse, General Electric, and so on, did their own research *and* funded what little research there was at the university level. It wasn't until after WWII that the US adopted the "Germanic" model of the research university, funded largely by DOD. Classified research was done at those universities until the late 60s, when the Gimmie Generation threw them off campus (forming SRI, MITRE and the rest). Funding by DoD continued strong into the early 1990s (DoD funded 2/3 of all computer science research in the late 80s), but then died with Al "Well, at least I funded the Internet" Gore looking on.
Perhaps you'd prefer that we went back to the British university model instead? Oddly enough, I agree somewhat, but then, where would we find anyone who still actually knows how to teach?
Having done my graduate work at a major research university, I'd have to say that such universities are really research institutes that run a school "on the side". If you're a grad student, that's great, because you'll wind up learning far more from your projects than you will in the classroom. But, you will be faced with time pressures to complete projects, no matter who the patron is -- government funded research also has deadlines, checkpoints, and progress reports to write.
As to intellectual property, whoever pays for the research should reap the rewards. We the People pay for government-funded research, which is why the public owns the intellectual property for those projects. But, when a for-profit company is the patron, you'd better believe that they have every right to the ideas, work product, and patents that come out of it.
In recent years, corporate funding has been used to supplant declining technology research funding from the federal government (oh yes, thank you Al Gore for axing the research budget). Thus the patent issue has arisen, mostly in engineering research. Frankly, if you don't like patents, don't take the research money. But then you can most likely forget about studying engineering. Go study poetry instead, which is (of course) funded by university indirect fees on research...
"Nearly everything has evolved in my lifetime except the antenna."
Well son, where the hell have you been? Electrically variable resonant circuits have been a feature of RF synthesizers since at least the late 70s. Phased arrays, scanning arrays all use similar techniques, as have electrically adjustable filters based on nonlinear materials such as indium antimonide. This is a nice approach, but it's hardly revoultionary and still doesn't solve the isolation problem or linearity issues for CDMA.
Look, SlashDot's staff is perfectly capable of reporting on software issues and leftist politics, and can probably write a good line of code. Stick to your expertise. Don't bother bullshitting us about RF if you can't hack vector calculus, don't dream Maxwell in your sleep, haven't brought a Beowulf cluster to its knees doing FEA, never drew an arc from a kilovolt power supply while warming yourself above those cherry red vacuum tubes in the final, and can't sling at least 20 WPM from a Vibroplex...
People often forget that the railroads ran the "first Internet", an interconnected network of telegraph lines, relays (routers) and stations (end nodes), used to send telegrams (email), over a century ago.
As for signalling, the railroads have used Centralized Traffic Control (CTC), a digital network technology, for almost 70 years. The basis of the system is a series of frequency coded pulses, which are sent over a separate pole line, but often repeated into the rails to be inductively picked up and displayed in the locomotive cab. CTC is used extensively on major mainlines in the US, and since 1980, CTC districts on the two major western US railroads have been controlled from a single dispatching center (the Union Pacific in Omaha and the Burlington Northern Santa Fe in Ft. Worth(less).
Well, we've pretty much gone our separate ways. Last I heard, Louis was off at Kleiner-Perkins. Mike is still at SRC. Me? I punched out of AV and went to Zindigo a few months ago.
Yes, I'm bummed about how things turned out, but The Show Goes On...
-=paulf
College is not about the first job you'll get out of school. Rather, it's a preparation for the rest of your life...for *all* of the jobs you'll have. Sure, right now, CIS might pay as well as CS, but that will probably change 5-10 years down the road. Moreover, *you* will change during the next 5-10 years. Maybe you'll want to do something else with your life -- but if you don't have a more general education, that will be impossible, you'll be trapped!
Math is never easy for anyone. For what it's worth, the math required for most CS degrees is much easier than for electrical or mechanical engineering. But math is the great societal dividing line -- it will determine what you can and can't do in your career like nothing else. Like a good workout, it's worth it in the end.
"Pain is weakness leaving the body. -- USMC"
Slashdot Mirror
Never discount the value of dusty decks and Code Inertia.
Frankly, nobody in their right fleeping mind does honest realtime work on Linux; if anything, the various tiny BSD kernels would be the tool of choice. Most realtime code these days is written either for nuxi-like environments like VxWorks, or hand bummed for custom threaded operating systems.
I wouldn't exactly proclaim ham radio as a bearer of tech progress either. As I stated above, the innovations have been more economic than technical at times. But give credit where credit is due. Where you around for the Sheet Metal Days of microcomputing? How do you think the hardware community communicated and swapped fixes back then? Yes, you could wait for the bi-monthly flimsy newsletter that some poor hack put out on an irregular basis. Or you could instead interact with the weekly net on 20 meters, ask if anyone had seen a problem or had a fix, and have them describe it to you directly. The choice is pretty obvious.
As time wore on, programs written for the x80-CP/M environment were ported to the x86-DOS world because that's what was available, and most radio applications required what the PC gave you -- the whole machine. But, in all this time, there was never a heritage of code-sharing, since making a little money on the side allowed you to defray the costs of the hobby. That's one well-known characteristic of hams -- they're cheapskates by nature or pick it up as they go on -- and as a result many ham radio innovations are economic, not technical.
As more modern multitasking operating systems showed up, non-realtime apps have been ported to them, and some amazing semi-realtime DSP work has been done, such as the various PSK31 implementations. And many of these are at least of an Open philosophy -- almost all of the PSK31 implementations, for example, are based on a single core DLL produced under a "share and enjoy" license.
So indeed, amateur radio comes from the DOS world, and the Free Nuxis have a lot of catching up to do.
1. Internal. This is the Winradio approach. The good news is that, as processors become faster and faster, they're able to absorb more of the electronics into software layers. The latest WinRadio is akin to all those Winmodems we've seen. However, the environment within a computer case isn't exactly the best place to put an RF circuit -- it's full of all kinds of strong fields and oddball harmonics.
2. External. A much nicer place to put your radio is in a nice RF tight box a few feet away from all that nonlinear harmonic crapola. And, after all, the output is relatively low bandwidth, so bring it into the system through an I/O port -- USB, 1394, heck even a serial port will do.
What you really want is an Icom PCR-1000, covers 100 kHz to 1.3 GHz (continuous if you shop in Akihabara), multimode. Hook it up to a serial port and an audio in jack, and you're all set to vacuum the ether.
Or, just check out the JavaRadio network of PCR-1000 equipped sites around the world...
I'm using an Eden 5000 right now for my DSP work. Together with VIA's low power chipset, you can build a fanless system for less than $400.
...the last remaining religion that atheism can bash with impunity. Anything else would be antisemitic, or a hate crime. Even Islam sees more politically correct protection these days ("mustn't incite the people to war, you know...").
There are good, sound, moral reasons for opposing embryonic stem cell research, if you have any respect for life or the living that a belief in a Deity has nothing to do with. But, rather than address the arguments logically, Reeves has instead chosen the Ad Hominem. Swell.
While Intel and AMD have been locked in the performance race, Cyrix took their x86 line in a different direction: low power consumption and low heat dissipation. I just built a new box based on the VIA/Cyrix Eden ITX formfactor motherboard for less than $400, and it consumes less than 36w (no fan). And, best of all, it runs everything in the ham shack -- including some powerful DSP software (PSK31).
Yes, but let's face it: none of these shows holds a candle to the true King of Implied Violence Entertainment: Jerry Springer. Thing is, if you tune into Springer, there's an expectation of (at least thinly veiled) violence. As far as I'm able to tell, Phil Donohue hasn't hired bouncers...yet.
But then again, there's always been a certain thrill to heated arguments. Anybody remember Point / Counterpoint on 60 Minutes? [Okay, we *all* remember the SNL parody...] So this is nothing new.
Oops, sorry, wrong Alpha...
Right now, Caltrain is running at Restricted Speed south of Redwood Junction, I'm listening to the dispatcher dish out track-n-time as the track and bridges get inspected...
I sincerely doubt, for example, that the author of this study took into account motion of the "point sources". Next time you ride a train, look at how people tend to rock back and forth as the suspension dampens the shock of rail joints and other track discontinuities. It's not much, maybe five to ten centimeters, but that's enough to cancel out constructive interference when the wavelength is only about 16cm.
This is why experimental science and empirical engineering are so crucial; they either confirm our understanding of the problem, or it's back the drawing board....
Yes, I would wonder that, if there were any truth to it. My Motorola CDMA Vader is just as small as the GSM Vader. And I have an 802.11b card (Symbol's CF Wi-Fi) that's smaller than any GSM phone. Don't overgeneralize.
When it comes down to it, Spread Spectrum is nothing more than a spectrum allocation algorithm that depends on the orthogonality of the chipping sequences. This orthogonality is further degraded by Rayleigh fading (aka multipathing), which is why ACPR or diversity reception is essential to CDMA. The primary benefit, however, is that that as an allocation scheme, spread spectrum is time based, and can take advantage of the bursty nature of most communications, be they human-to-human or machine-machine. And that's why time domain approaches are more efficient.
The amateur service is intended to be more than merely a backup yacking system to cellphones. It's supposed to encourage experimentation with new systems. Sadly, the current FCC rules don't allow for much useful spread spectrum experimentation by hams, but the truth of the matter is that existing commercial systems could be ported to the amateur frequencies, and do the jobs you mention (emergency communications) far, far better than the 1950s FM technology you're clinging to.
Look at what happened to Packet Radio in the Amateur Service; we built it, they came, but then the FCC issued and enforced inane and archaic content restrictions on the use of packet, and the experimentation died out. The FCC is doing it to us again, all in the name of control.
Cory Doctrow's article, BTW, is riddled with errors and false assumptions about the ranking algorithms used by both Google and AltaVista, which have evolved radically since their introduction. AltaVista, for example, ceased to be a purely keyword-based engine even before it hit the street. And, now that Google actually has a substantially sized audience, they're having to deal with a lot more spam attacks than ever before.
The real battle to come isn't going to be entirely technology based, however. It's going to be a duel between CMGI and Kleiner-Perkins to see who can figure out how to make a profitable (with good ROIC) search engine business before AltaVista and Google drain their coffers. In that aspect, Google is woefully behind.
-=paulf
If you're 26 years old, then you're too young to remember the Cold War. You know, the one we WON because nobody in their freaking mind would attack the US military? Well, guess what: eight years of military softening by the Philanderer-in-Chief left not only our military but our CIVILIAN population vulnerable to some psycho with a Weapon of Mass Destruction (and yes, a 767 at speed loaded with fuel qualifies as a WMD).
There are only two options for dealing with such people. One is to eliminate any civil liberties, and spot them as they're about to attack. The other is to kill them in place. I'll opt for the "Kill in Place" method, since I see no reason to screw up a perfectly good society.
As for the rest of the world, they do what's in their best interest without consulting the US. Get a clue.
-=paulf
No, the biggest problem with 802.11a is that it isn't spread spectrum. Yes, I know that some adherents like to think that FH is SS, but a quick look at a spectrum analyzer will show you how unspread the signal is, full of peaks and impulses. And then there's the 5 GHz question. Four times the path loss, and unwanted gain.
Face it, backward compatibility matters. And for the broader market, that means 2.4 GHz and something that will do "b" in a pinch. One can argue technical merits until you're blue in the face, but what really matters is the market, and if the history of POTS modem development is any guide, "g" is the future.
The first spinout attempt was lead by a CEO who wanted AltaVista to be a software company, and was ended by the 1997 market plunge. Then Compaq bought DEC, and eventually sold most of AltaVista to CMGI. They then attempted to convert AltaVista into a portal, since the investment-driven strategy of the day called for eyeballs and land grab. A huge amount of cash was spent on the portal strategy (and very little on the search engine) until the day in April 2000 when eyeball valuations died.
The moral of the story is that market and capital strategy matter just as much as technology.
How do I know? I was one of those three guys!
Many businesses are now failing, not because of bad technology or bad business models, but because of poor execution. Startups are particularly beholden to their early investors, and all of the VCs were heavily pushing "land grab" business practices throughout the bubble. The reason is obvious: Stocks were valued primarily on audience, and if you want the best quick return on your investment, you demand practices that build audience share, period. Early investors don't care about long term prospects for a business, they just want a quick cash-out to flip into something else. The 1997 changes in post IPO-lockout (from two years to six months) only magnified the short term focus. Having structured the business with that bias, the result is inevitable.
Businesses fail for all kinds of reasons, many of which are completely out of control of the management or techs. Don't dismiss an idea just because somebody screwed the pooch in the past.
Alpha was the great architecture that blew everything else away, but thanks to the guys in Houston, we're left with the One True Architecture. Bah!
"The 80xxx series of microprocessors is clear evidence that INTEL isn't doing in-house drug testing." (Usenet, 1988)
"It would be different if Slashdot made any claims to political, social, or cultural neutrality, but the fact is that it doesn't. If you read this site, you should expect a liberal bent. If you want a rightwing take on the same issues, I'm sure Google will dig up plenty of suitable articles and discussions. "
Spoken like a true soldier of Political Correctness. "If you're in the University, you should expect a liberal slant, and if you don't like it, go someplace else!". Yeah, right. Slashdot proclaims itself as News for Nerds. Last time I checked, Engineers were for the most part considered Nerds, and more than three out of four Engineers consider themselves politically conservative. You'll find just the opposite numbers for Computer Scientists (and leaning but not quite as far left for most of the physical sciences). Slashdot, for all intents and purposes, is News for CS Types , and the resulting political bent comes as something of a shock for those who haven't figured out that Computer Scientists are not Engineers!
Actually, up until about 1930 or so, most innovation in the US *was* privately funded -- because the federal government lacked the money before the passage of the income tax as a method of confiscating wealth. Companies like RCA, Philips, US Steel, Westinghouse, General Electric, and so on, did their own research *and* funded what little research there was at the university level. It wasn't until after WWII that the US adopted the "Germanic" model of the research university, funded largely by DOD. Classified research was done at those universities until the late 60s, when the Gimmie Generation threw them off campus (forming SRI, MITRE and the rest). Funding by DoD continued strong into the early 1990s (DoD funded 2/3 of all computer science research in the late 80s), but then died with Al "Well, at least I funded the Internet" Gore looking on.
Perhaps you'd prefer that we went back to the British university model instead? Oddly enough, I agree somewhat, but then, where would we find anyone who still actually knows how to teach?
Having done my graduate work at a major research university, I'd have to say that such universities are really research institutes that run a school "on the side". If you're a grad student, that's great, because you'll wind up learning far more from your projects than you will in the classroom. But, you will be faced with time pressures to complete projects, no matter who the patron is -- government funded research also has deadlines, checkpoints, and progress reports to write.
As to intellectual property, whoever pays for the research should reap the rewards. We the People pay for government-funded research, which is why the public owns the intellectual property for those projects. But, when a for-profit company is the patron, you'd better believe that they have every right to the ideas, work product, and patents that come out of it.
In recent years, corporate funding has been used to supplant declining technology research funding from the federal government (oh yes, thank you Al Gore for axing the research budget). Thus the patent issue has arisen, mostly in engineering research. Frankly, if you don't like patents, don't take the research money. But then you can most likely forget about studying engineering. Go study poetry instead, which is (of course) funded by university indirect fees on research...
Well son, where the hell have you been? Electrically variable resonant circuits have been a feature of RF synthesizers since at least the late 70s. Phased arrays, scanning arrays all use similar techniques, as have electrically adjustable filters based on nonlinear materials such as indium antimonide. This is a nice approach, but it's hardly revoultionary and still doesn't solve the isolation problem or linearity issues for CDMA.
Look, SlashDot's staff is perfectly capable of reporting on software issues and leftist politics, and can probably write a good line of code. Stick to your expertise. Don't bother bullshitting us about RF if you can't hack vector calculus, don't dream Maxwell in your sleep, haven't brought a Beowulf cluster to its knees doing FEA, never drew an arc from a kilovolt power supply while warming yourself above those cherry red vacuum tubes in the final, and can't sling at least 20 WPM from a Vibroplex...
People often forget that the railroads ran the "first Internet", an interconnected network of telegraph lines, relays (routers) and stations (end nodes), used to send telegrams (email), over a century ago.
As for signalling, the railroads have used Centralized Traffic Control (CTC), a digital network technology, for almost 70 years. The basis of the system is a series of frequency coded pulses, which are sent over a separate pole line, but often repeated into the rails to be inductively picked up and displayed in the locomotive cab. CTC is used extensively on major mainlines in the US, and since 1980, CTC districts on the two major western US railroads have been controlled from a single dispatching center (the Union Pacific in Omaha and the Burlington Northern Santa Fe in Ft. Worth(less).
Well, we've pretty much gone our separate ways. Last I heard, Louis was off at Kleiner-Perkins. Mike is still at SRC. Me? I punched out of AV and went to Zindigo a few months ago. Yes, I'm bummed about how things turned out, but The Show Goes On...
-=paulf
College is not about the first job you'll get out of school. Rather, it's a preparation for the rest of your life...for *all* of the jobs you'll have. Sure, right now, CIS might pay as well as CS, but that will probably change 5-10 years down the road. Moreover, *you* will change during the next 5-10 years. Maybe you'll want to do something else with your life -- but if you don't have a more general education, that will be impossible, you'll be trapped! Math is never easy for anyone. For what it's worth, the math required for most CS degrees is much easier than for electrical or mechanical engineering. But math is the great societal dividing line -- it will determine what you can and can't do in your career like nothing else. Like a good workout, it's worth it in the end. "Pain is weakness leaving the body. -- USMC"