"This 'telephone' has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us." --Western Union internal memo, 1876
Lots of folks have rightly pointed out that the electronics for cellular telephones wasn't ready in 1947 but neither was the switching infrastructure. The first customer-dialed long distance telephone call wasn't placed until 1951 and it took at least the better part of a decade for that technology to be widely available across the US. Telephone switching was largely electromechanical until the early 1970s with the first Electronic Switching System having been deployed in 1965. But even at that time, the #1 ESS didn't have the sophistication or compute power to keep track of phones and perform handoffs as they moved from cell to cell. (The first true cell phone networks were built on the 1A ESS that came over 10 years later.)
Having said that, AT&T did demonstrate the cellular concept with a pay phone service on the Metroliner train service between New York City and Washington DC in the late 1960s. The pay phones used land mobile radio channels and would switch from channel to channel with a reuse pattern up and down the Northeast Corridor. There was a lot of work that had to be done and technology that had to be developed to get to that point - it took them a good 20 years to build this proof of concept, which they did prior to the FCC allocating what ultimately would become cellular spectrum. In fact, the Metroliner payphone system was a key milestone in opening up the conversation to get UHF TV channels 70-83 allocated for cellular.
TFA would have been more accurate if it had focused on 1970s as lost time for cellular development, but I think it was only a marginal effect. Much of the technology was still being developed (like Marty Cooper's handheld cell phone) while the lawyers and lobbyists haggled over the spectrum. By the point, it was pretty clear cellular was going to happen, it was just the regulatory details that had to be worked out.
There were a number of Sun workstations inside the CM5 that served as partition managers, but the computational horsepower was custom TMC hardware based on the SPARC processor. The CM5 was much more than a rack of Sun workstations. At NCSA, two of the cabinets had the Sun (SPARCstation 10???) partition managers, but other three cabinets had nothing but the TMC hardware in them.
Actually, in the 19th century they did first come up with a data network. It was called the telegraph and used Morse code. It was only later that they figured out how to put voice over the wires.
And the dominant telegraph company wanted nothing to do with telephone at the time. As soon as I read the above posting I immediately thought of the quote:
This 'telephone' has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us. -- Western Union internal memo, 1876.
Hmmm... An entenched company resisting new technology... Now that sounds vaguely familliar.
Only after Bell started selling service did WU change their tune and try to pursue the business.
From the 1950s to 1970s programming was considered a trade school discipline. MIT avoided even offering a major in the subject.
Does anyone else remember the TV adds for "Control Data Institute?" I rember seeing them watching afternoon TV as a kid in the 70s. (CDI adds ran between the "Meet Chuck" mechanic school ads and the DeVry ads.) CDI was on offshoot of Control Data to teach programmers in a trade school environment.
The plumber/electrician analogy is very apt. You wouldn't hire an electrical engineer to do the work of an electrician. An engineer may understand and specify an electrical system on paper, but it takes the equally important skill set of the electrican to get installed efficiently and properly. The problem is that many employers inappropriately focus on BSCS degrees for all IT jobs, probably because there are a lack of real quality "programming" curricula out there. (As a side note, while I'd probably agree some of the best IT people pick up the knowledge on their own without getting a technology degree, I would argue that there may not be enough of that type of people around...)
Re:Yep, RoadRailers are nothing new.
on
By Road and Rail?
·
· Score: 1
The technology, believe it or not, is close to 50 years old. The Chesapeake and Ohio Historical Society has photos of RoadRailers in revenue service in 1960! It's ironic that Norfolk Southern is the champion of the service today whereas the C&O was a major proponent and developer of the technology. The C&O, of course, was a predecessor of CSX, Norfolk Southern's chief competitor.
The quality of service at the Residential end should be dropping remarkably over the next few months, combined with a lack of solid information for reps and customers alike.
Good gosh, their customer service can get worse?!?! I jumped ship from AT&T about 6 months ago after a horrible experience with changing plans. The customer service folks I worked with were either completely incompetent or just couldn't care less about my business. Every phone call to them (which dragged over two months time) was a time-wasting ordeal. There was no continuity and I had to start from scratch every time I called. I had been a customer of theirs for 15 years and they did everything they could do to drive me away.
I'm sorry, but there's a difference between supporting many serial ports (or network terminals) and supporting multiple video adaptors. Linux had always been able to work as a terminal server. This is something new.
I'm not sure this is new. I believe the Stardent Titan Graphic Supercomputer I worked on in 1990 had the capability of supporting two graphics consoles (FBs, keyboards, mice, etc.) with two independent X servers running. Our system at Purdue only had one graphics head (and it was pretty buggy at that) but I seem to remeber learning that that the "a" in "hostname:a.b" was for selecting the particular graphics terminal (and corresponding X server) while reading the documentation for that beast. IIRC, the actual console (i.e. the terminal you typed on to boot it up) for that machine was a serial-tty off a supervisor/network board next to the R3000 CPU boards while the graphics board(s) sat in one (or both) end(s) of the card cage. These graphic systems were peripheral systems which were brought up after the machine was booted via the serial console. The idea being the system could be purchased with 0, 1, or 2 graphics head to suit the users' needs. (For the day, these were very expensive grpahics systems.) I believe SGI systems of the day or very shortly after were available with the same capability.
In any event, I really don't see what the big deal is. In the '80s there were numerous universities running UNIX systems that simultaneously supported hundreds of concurrent users. Yeah, they were doing it with serial (dumb) terminals or (semi-intelligent) X terminals, but the end result was largely the same. I wonder if a Linux/FreeBSD box could really support 150 undergrads doing their homework like the {super}minicomputers of that era could. (Imagine typing who and watching 150 logins scroll by. You almost always had to pipe who or finger into more!) Tuning BSD 4.x and its variants to operate under that load was quite an accomplishement.
Ditto on the Tek 2235 recomendation, or any 2200 series 'scope for that matter. I learned on a 2235 in high school. (Of course it was one of the newest portable 'scopes available at that time...) I've used a lot of different scopes, both analog and digital, and the 2235 is still hands down my favorite. It's a great general-purpose instrument that is easy to use, perfect to learn and grow on.
I know the original poster asked for a PC-based instument and he may have a good reason, but I obviously fall into the non-PC-based crowd. Good used scopes are available at cheap prices, there's no reason the kids can't use the real thing. Yes, PC-based instruments have some unique capabilities, but most of those features are overkill or a distraction in a high school classroom.
Oh the Empire Carpet guy... But he's still on the air (I think) so probably the most remembered in Chicago... Do you remember the phone numbers for:
1. First Metropolitan Builders
2. Bouchelle Carpet Cleaning
3. Lincoln Carpeting
If you know those numbers, you are a true, old-school Chicago television junkie.
The answers are:
1. 282-8600. No fancy jingles or gimmicks; these guys beat the number into you with sheer repetition...
2. In your best, deep bass voice: Hudson-three-two-seven-hun-drehhhhhhhhhhhd (HUdson3-2700)
3. C'mon, sing the jingle: When you're thinkin'/ Lincoln Lincoln/ Better carpeting for less / call National-two-nine-thou-ou-sand / National-two-nine [telephone ring] thousand. (NAtional2-9000)
--zawada (Who spent way too much time watching television as a kid in the '70s and '80s)
Apparent power, measured in Volt-Amperes (VA), is the vector sum of real power, measured in Watts(W), and reactive power, measured in Volt-Amperes-reactive (VAr). When the power factor = 1.0, MVA=MW. Thus MW isn't the answer either.
> Well, I have news for you: 50MV lines don't exist! Not out in the open, anyway. Was it 50 kV, perchance?
>>nope, MV... though it may have been 45MV...
The first guy is right; there is no such thing as a 45 MV transmission line. The highest voltage transmission line classification is 765 kV. (That would be 0.765 MV.) In the mid-1970s American Electric Power and Ohio Brass played with some experimental 1.5 MV transmission equipment but they killed the project when they realized land owners would never let AEP put a 1.5MV line in their back yards.
The lines that First Energy put in the trees were 345 kV. I'm guessing they were rated to carry between 1000 to 1500 MVA. I have no idea where the 45 number came from or what unit would have been associated with it.
The other humbling thing I learned was that if it wasn't on the negative, it wasn't going to be in the print. You can't coax a masterpiece from a mediocre piece of film.
This is very true. Adams employed the Zone system throughout the photographic process from exposure and developing of the film to printing on paper. He published a great book called Examples: The Making of 40 Photographs which goes into great detail on how each of 40 of his most well-known images were produced. In many cases he describes how he assigned Zone values to various elements of the photograph at the time of exposure and carried them through the whole production of the image.
Adams planned his shots, set up his big old camera, then waited for the scene to appear and the light to be just right. Click! You only get one chance when you do it this way because it might take half a day to prepare. His negatives were awesome!
Yes but not always, in the aforementioned book I believe he described photographing Moonrise over Hernandez, New Mexico in quite a hurry after seeing the image set itself up while driving close to dusk. He knew his technology so well that he was able to apply the Zone principles quickly and still get the shot. So while he did plan many of his shots, he could on occasion think on his feet and get a good negative quickly.
Though it's a parody and I generally try to take those lightly, he's made one critical error that really stands out in his assertion that free software is the domain of hackers/tinkerers/students, etc. I think Howard Strauss ought to be informed of the billions of dollars being invested in free software development by major corporations, many of whom have salaried and talented employees developing such applications. His condescending attitude towards the talented programmers who have created so much of the infrastructure the Internet depends on (Linux, BSD, Apache, MySQL anyone?) is a bit infuriating, to say the least.
Not only is he ignoring current developers on a major corporations' payrolls, but he also shows he's ignorant of at least some of the origins of the OSS movement. A lot of folks that developed OSS in the late '80s and early '90s were university-types who were developing code for production computing environments. Many of the early OSS packages were developed because they addressed areas that the marketplace left untouched, especially when it came to UNIX "academic computing" users. (As a side note, I don't think you see nearly as much good OSS coming out of universities anymore because most have become aware of what was going on and they've set up fences to hold on to worthwhile intelectual property in an attempt to make money off of it.)
It could also be that he's perfectly aware of where OSS came from and that this is an extension of an old rift between "academic computing" and "admininistrative computing". Howard Strauss represents administrative computing which stuck to traditional, commercial solutions like Big Blue mainframes. OSS represents the academic computing folks who have always been left to fend for themselves but typically do a damn good job with much fewer resources. Now that the academic-type solutions are vying for a place in the administrative domain, Strauss is circling the wagons to keep the ne'er-do-wells out.
"MSEE is on the city grid, not the university grid, though, because of its location and newness.
Are you sure about that? My recollection was the Campus was a grid unto itself with several interconnections to Cinergy/PSI around campus. Maybe my memory is failing though...
The MSEE building opened when I was an undergrad at Purdue and worked for ECN. Back then, we had a ton of "big machines" (Goulds and Vaxes) in the now-gone MSEE104 machine room as well at many of the other engineering building on campus. One reason we kept a radio scanner on the Physical Plant two-way radio system was to listen for (IIRC) "unit 519" who was the power plant turbine operator. If he started to report problems, people would scramble to prepare for the ensuing flaky power problems. That was back in the late 1980s and I think they've added more generating capacity since then, though I don't know if the campus power plant can carry the entire campus.
You're right though, MSEE has always seemed to have more than its share of power problems. The most spectacular that I recall was a couple years after the building was built. Moisture worked its way into some underground 12 kV splices and blew a manhole cover into the air between the EE and MSEE buildings... Luckily no one was hurt but it caused quite a disruption.
Your point is correct... It's not the mechanical and electrical systems that failed but the logistical systems that crippled the systems.
I don't know about Germany, but in the US the basic functionality and "interlocking" logic of railroad signalling systems is located trackside in the field and (like the parent poster says) is usually made up of the same old klunky mechanical devices the rail industry has used for the last 75 years. These devices may be controlled by microprocessors (and ulimately by some windows-based device at a dispatching center) but thier hardwired logic will not allow dangerous conditions like two opposing "clear" signals for the same piece of track. The industry has grafted information-age technology to the old system, but the tried-and-true failsafes are still there.
The reason CSX had to shut their rail system down was that their central dispatch people were "blind" and couldn't direct train movements. In railroading the dispatchers make most of the real-time operating decisions. And if the dispatchers can't keep track of all of their trains, they have a serious problem. A logistic nightmare yes- safety nightmare, not likely... No lives were lost but it was a serious PITA for those commuters.
"The problem is not the SCADA or braking system itself; it's the remote monitoring station."
Mod this guy up; he knows what he's talking about... If you read the CSX press release, this is what they are saying... The signal system itself didn't fail, it's all of the support systems that the dispatchers use that fell apart. And since it's the dispatchers who decide which train goes where (not the engineer and conductor in the locomotive cab) the CSX rail system became crippled when their network got trashed...
No: saying someone is wrong just because he doesn't have a background in that area without countering the claims or showing that the premise is incorrect is pseudoscience. Real science is talking about the actual claims, not about the person making them.
To some degree I agree with you, however, the Open Spectrum crowd is lead primarily by computer scientists who as far I can tell have ignored some of the issues that crop up when designing radio real radio systems, especially when it comes to receivers. They wave their hands and say this new paradigm will leave all the old technology behind.
All of the literature I have read has focused on interference as occurring when two narrow band transmitters try using the same channel. This is the obvious case and the reason we have distinct allocations of the electromagnetic spectrum today. However, there are other modes of interference that occur even when two transmitters are transmitting on different channels. What about intermodulation interference (intermod)? Intermod occurs when two signals mix (multiply by one another) in the non-linear portions of a receiver to produce spurious signals that wipe out the real signal of interest. What about desense which occurs when one strong signal kilohertz or megahertz away from the signal of interest swamps the front end and makes the receiver deaf to all other signals? These types of problems have been dealt with by RF engineers for years in an environment where the spectrum was supposedly cafefully managed. (Some would say micromanaged.) If the spectrum turns into a free-for-all, I shudder to think the problems that will arise. Transmitting at a low power will mitigate these problems to a degree, but not all radio applications can be low power. (Think rural areas...)
You want to talk about actual claims? How about the claim that software defined radios (SDRs) are frequency agile and can operate anywhere they want. What you will find if you actually try to build an SDR receiver is that current CPU and A-to-D technology is not fast enough to work directly in the E-M spectrum. Depending on the design, you'll be limited to 0-30 MHz at best. To go higher, you'll need to use one of two methods, downconversion (mixing) or undersampling. Both of these processes require the use of filters and possibly other analog components that bring many of the same problems as they do in traditional radios. There are similar issues on the transmit side. There are still some excellent reasons to do SDR instead of traditional radios, however true frequency agility is still far out on the horizon.
I like many aspects of the Open Spectrum movement. There are some workable concepts there. However, I think it's being overtaken with hype and I worry that hype will hurt the credibility of the truly valid concepts. When I see engineers with real RF experience making the Open Spectrum movement's claims, I will be much less skeptical.
If you want to use HF, you're probably going to need it because the channel allocations are so small.
No, the FCC requires you to know code to legally operate on HF because an international treaty requires them to. I believe there is a proposal on the table at WRC-03 to remove that requirement.
Slashdot Mirror
I call BS on any survey or programmer that considers HTML a programming language.
"This 'telephone' has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us." --Western Union internal memo, 1876
Lots of folks have rightly pointed out that the electronics for cellular telephones wasn't ready in 1947 but neither was the switching infrastructure. The first customer-dialed long distance telephone call wasn't placed until 1951 and it took at least the better part of a decade for that technology to be widely available across the US. Telephone switching was largely electromechanical until the early 1970s with the first Electronic Switching System having been deployed in 1965. But even at that time, the #1 ESS didn't have the sophistication or compute power to keep track of phones and perform handoffs as they moved from cell to cell. (The first true cell phone networks were built on the 1A ESS that came over 10 years later.) Having said that, AT&T did demonstrate the cellular concept with a pay phone service on the Metroliner train service between New York City and Washington DC in the late 1960s. The pay phones used land mobile radio channels and would switch from channel to channel with a reuse pattern up and down the Northeast Corridor. There was a lot of work that had to be done and technology that had to be developed to get to that point - it took them a good 20 years to build this proof of concept, which they did prior to the FCC allocating what ultimately would become cellular spectrum. In fact, the Metroliner payphone system was a key milestone in opening up the conversation to get UHF TV channels 70-83 allocated for cellular. TFA would have been more accurate if it had focused on 1970s as lost time for cellular development, but I think it was only a marginal effect. Much of the technology was still being developed (like Marty Cooper's handheld cell phone) while the lawyers and lobbyists haggled over the spectrum. By the point, it was pretty clear cellular was going to happen, it was just the regulatory details that had to be worked out.
Yeah, and those would be the same monkeys that might fly out of my butt!
What no ATM network? (That would be Asynchronous Transfer Mode, not any kind of banking machine)
A dozen years ago the "experts" were predicting that the LAN would be replaced by an ATM WAN...
--z (who suffered the pain of implementing "LAN Emulation" on ATM)
There were a number of Sun workstations inside the CM5 that served as partition managers, but the computational horsepower was custom TMC hardware based on the SPARC processor. The CM5 was much more than a rack of Sun workstations. At NCSA, two of the cabinets had the Sun (SPARCstation 10???) partition managers, but other three cabinets had nothing but the TMC hardware in them.
And the dominant telegraph company wanted nothing to do with telephone at the time. As soon as I read the above posting I immediately thought of the quote:
This 'telephone' has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us. -- Western Union internal memo, 1876.
Hmmm... An entenched company resisting new technology... Now that sounds vaguely familliar. Only after Bell started selling service did WU change their tune and try to pursue the business.
--zawada
Ummm... Wasn't that Comcast that wanted to buy Disney?
Does anyone else remember the TV adds for "Control Data Institute?" I rember seeing them watching afternoon TV as a kid in the 70s. (CDI adds ran between the "Meet Chuck" mechanic school ads and the DeVry ads.) CDI was on offshoot of Control Data to teach programmers in a trade school environment.
The plumber/electrician analogy is very apt. You wouldn't hire an electrical engineer to do the work of an electrician. An engineer may understand and specify an electrical system on paper, but it takes the equally important skill set of the electrican to get installed efficiently and properly. The problem is that many employers inappropriately focus on BSCS degrees for all IT jobs, probably because there are a lack of real quality "programming" curricula out there. (As a side note, while I'd probably agree some of the best IT people pick up the knowledge on their own without getting a technology degree, I would argue that there may not be enough of that type of people around...)
My alma matter, Purdue has addressed this situation with two programs. One is a Conventional BSCS program in the School of Science, the other is a rigerous Computer Technology program in the School of Technology.
--zawada
The technology, believe it or not, is close to 50 years old. The Chesapeake and Ohio Historical Society has photos of RoadRailers in revenue service in 1960! It's ironic that Norfolk Southern is the champion of the service today whereas the C&O was a major proponent and developer of the technology. The C&O, of course, was a predecessor of CSX, Norfolk Southern's chief competitor.
--zawada
Good gosh, their customer service can get worse?!?! I jumped ship from AT&T about 6 months ago after a horrible experience with changing plans. The customer service folks I worked with were either completely incompetent or just couldn't care less about my business. Every phone call to them (which dragged over two months time) was a time-wasting ordeal. There was no continuity and I had to start from scratch every time I called. I had been a customer of theirs for 15 years and they did everything they could do to drive me away.
--zawada
I'm not sure this is new. I believe the Stardent Titan Graphic Supercomputer I worked on in 1990 had the capability of supporting two graphics consoles (FBs, keyboards, mice, etc.) with two independent X servers running. Our system at Purdue only had one graphics head (and it was pretty buggy at that) but I seem to remeber learning that that the "a" in "hostname:a.b" was for selecting the particular graphics terminal (and corresponding X server) while reading the documentation for that beast. IIRC, the actual console (i.e. the terminal you typed on to boot it up) for that machine was a serial-tty off a supervisor/network board next to the R3000 CPU boards while the graphics board(s) sat in one (or both) end(s) of the card cage. These graphic systems were peripheral systems which were brought up after the machine was booted via the serial console. The idea being the system could be purchased with 0, 1, or 2 graphics head to suit the users' needs. (For the day, these were very expensive grpahics systems.) I believe SGI systems of the day or very shortly after were available with the same capability.
In any event, I really don't see what the big deal is. In the '80s there were numerous universities running UNIX systems that simultaneously supported hundreds of concurrent users. Yeah, they were doing it with serial (dumb) terminals or (semi-intelligent) X terminals, but the end result was largely the same. I wonder if a Linux/FreeBSD box could really support 150 undergrads doing their homework like the {super}minicomputers of that era could. (Imagine typing who and watching 150 logins scroll by. You almost always had to pipe who or finger into more!) Tuning BSD 4.x and its variants to operate under that load was quite an accomplishement.
--zawada
I know the original poster asked for a PC-based instument and he may have a good reason, but I obviously fall into the non-PC-based crowd. Good used scopes are available at cheap prices, there's no reason the kids can't use the real thing. Yes, PC-based instruments have some unique capabilities, but most of those features are overkill or a distraction in a high school classroom.
--zawada
1. First Metropolitan Builders
2. Bouchelle Carpet Cleaning
3. Lincoln Carpeting
If you know those numbers, you are a true, old-school Chicago television junkie.
The answers are:
1. 282-8600. No fancy jingles or gimmicks; these guys beat the number into you with sheer repetition...
2. In your best, deep bass voice: Hudson-three-two-seven-hun-drehhhhhhhhhhhd (HUdson3-2700)
3. C'mon, sing the jingle: When you're thinkin'/ Lincoln Lincoln/ Better carpeting for less / call National-two-nine-thou-ou-sand / National-two-nine [telephone ring] thousand. (NAtional2-9000)
--zawada (Who spent way too much time watching television as a kid in the '70s and '80s)
Apparent power, measured in Volt-Amperes (VA), is the vector sum of real power, measured in Watts(W), and reactive power, measured in Volt-Amperes-reactive (VAr). When the power factor = 1.0, MVA=MW. Thus MW isn't the answer either.
--zawada
>>nope, MV... though it may have been 45MV...
The first guy is right; there is no such thing as a 45 MV transmission line. The highest voltage transmission line classification is 765 kV. (That would be 0.765 MV.) In the mid-1970s American Electric Power and Ohio Brass played with some experimental 1.5 MV transmission equipment but they killed the project when they realized land owners would never let AEP put a 1.5MV line in their back yards.
The lines that First Energy put in the trees were 345 kV. I'm guessing they were rated to carry between 1000 to 1500 MVA. I have no idea where the 45 number came from or what unit would have been associated with it.
--zawada
This is very true. Adams employed the Zone system throughout the photographic process from exposure and developing of the film to printing on paper. He published a great book called Examples: The Making of 40 Photographs which goes into great detail on how each of 40 of his most well-known images were produced. In many cases he describes how he assigned Zone values to various elements of the photograph at the time of exposure and carried them through the whole production of the image.
Adams planned his shots, set up his big old camera, then waited for the scene to appear and the light to be just right. Click! You only get one chance when you do it this way because it might take half a day to prepare. His negatives were awesome!
Yes but not always, in the aforementioned book I believe he described photographing Moonrise over Hernandez, New Mexico in quite a hurry after seeing the image set itself up while driving close to dusk. He knew his technology so well that he was able to apply the Zone principles quickly and still get the shot. So while he did plan many of his shots, he could on occasion think on his feet and get a good negative quickly.
--zawada
Uh oh.. Looks like a memory hole swallowed www.thememoryhole.org:
;; res options: init recurs defnam dnsrch
;; got answer:
;; ->>HEADER ;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0
;; QUERY SECTION:
;; www.thememoryhole.org, type = NS, class = IN
;; Total query time: 6 msec
;; FROM: bliga.ncsa.uiuc.edu to SERVER: default -- 141.142.2.2
;; WHEN: Tue Nov 11 10:04:11 2003
;; MSG SIZE sent: 39 rcvd: 39
.org whois server though... Either a horribly ironic failure occured or the Memory Hole really exists... What gives?
% dig www.thememoryhole.org ns
; > DiG 8.3 > www.thememoryhole.org ns
It still showing up on the
--zawada
Not only is he ignoring current developers on a major corporations' payrolls, but he also shows he's ignorant of at least some of the origins of the OSS movement. A lot of folks that developed OSS in the late '80s and early '90s were university-types who were developing code for production computing environments. Many of the early OSS packages were developed because they addressed areas that the marketplace left untouched, especially when it came to UNIX "academic computing" users. (As a side note, I don't think you see nearly as much good OSS coming out of universities anymore because most have become aware of what was going on and they've set up fences to hold on to worthwhile intelectual property in an attempt to make money off of it.)
It could also be that he's perfectly aware of where OSS came from and that this is an extension of an old rift between "academic computing" and "admininistrative computing". Howard Strauss represents administrative computing which stuck to traditional, commercial solutions like Big Blue mainframes. OSS represents the academic computing folks who have always been left to fend for themselves but typically do a damn good job with much fewer resources. Now that the academic-type solutions are vying for a place in the administrative domain, Strauss is circling the wagons to keep the ne'er-do-wells out.
--zawada
Are you sure about that? My recollection was the Campus was a grid unto itself with several interconnections to Cinergy/PSI around campus. Maybe my memory is failing though...
The MSEE building opened when I was an undergrad at Purdue and worked for ECN. Back then, we had a ton of "big machines" (Goulds and Vaxes) in the now-gone MSEE104 machine room as well at many of the other engineering building on campus. One reason we kept a radio scanner on the Physical Plant two-way radio system was to listen for (IIRC) "unit 519" who was the power plant turbine operator. If he started to report problems, people would scramble to prepare for the ensuing flaky power problems. That was back in the late 1980s and I think they've added more generating capacity since then, though I don't know if the campus power plant can carry the entire campus.
You're right though, MSEE has always seemed to have more than its share of power problems. The most spectacular that I recall was a couple years after the building was built. Moisture worked its way into some underground 12 kV splices and blew a manhole cover into the air between the EE and MSEE buildings... Luckily no one was hurt but it caused quite a disruption.
--zawada
I don't know about Germany, but in the US the basic functionality and "interlocking" logic of railroad signalling systems is located trackside in the field and (like the parent poster says) is usually made up of the same old klunky mechanical devices the rail industry has used for the last 75 years. These devices may be controlled by microprocessors (and ulimately by some windows-based device at a dispatching center) but thier hardwired logic will not allow dangerous conditions like two opposing "clear" signals for the same piece of track. The industry has grafted information-age technology to the old system, but the tried-and-true failsafes are still there.
The reason CSX had to shut their rail system down was that their central dispatch people were "blind" and couldn't direct train movements. In railroading the dispatchers make most of the real-time operating decisions. And if the dispatchers can't keep track of all of their trains, they have a serious problem. A logistic nightmare yes- safety nightmare, not likely... No lives were lost but it was a serious PITA for those commuters.
--zawada
Mod this guy up; he knows what he's talking about... If you read the CSX press release, this is what they are saying... The signal system itself didn't fail, it's all of the support systems that the dispatchers use that fell apart. And since it's the dispatchers who decide which train goes where (not the engineer and conductor in the locomotive cab) the CSX rail system became crippled when their network got trashed...
--zawada
SCO in their best Dr. Evil voice:
"That's a number. Okay then. We hold the world ransom for.....One hundred..BILLION DOLLARS!!"
Hmm Hmm Ha Ha Ha!!!
To some degree I agree with you, however, the Open Spectrum crowd is lead primarily by computer scientists who as far I can tell have ignored some of the issues that crop up when designing radio real radio systems, especially when it comes to receivers. They wave their hands and say this new paradigm will leave all the old technology behind.
All of the literature I have read has focused on interference as occurring when two narrow band transmitters try using the same channel. This is the obvious case and the reason we have distinct allocations of the electromagnetic spectrum today. However, there are other modes of interference that occur even when two transmitters are transmitting on different channels. What about intermodulation interference (intermod)? Intermod occurs when two signals mix (multiply by one another) in the non-linear portions of a receiver to produce spurious signals that wipe out the real signal of interest. What about desense which occurs when one strong signal kilohertz or megahertz away from the signal of interest swamps the front end and makes the receiver deaf to all other signals? These types of problems have been dealt with by RF engineers for years in an environment where the spectrum was supposedly cafefully managed. (Some would say micromanaged.) If the spectrum turns into a free-for-all, I shudder to think the problems that will arise. Transmitting at a low power will mitigate these problems to a degree, but not all radio applications can be low power. (Think rural areas...)
You want to talk about actual claims? How about the claim that software defined radios (SDRs) are frequency agile and can operate anywhere they want. What you will find if you actually try to build an SDR receiver is that current CPU and A-to-D technology is not fast enough to work directly in the E-M spectrum. Depending on the design, you'll be limited to 0-30 MHz at best. To go higher, you'll need to use one of two methods, downconversion (mixing) or undersampling. Both of these processes require the use of filters and possibly other analog components that bring many of the same problems as they do in traditional radios. There are similar issues on the transmit side. There are still some excellent reasons to do SDR instead of traditional radios, however true frequency agility is still far out on the horizon.
I like many aspects of the Open Spectrum movement. There are some workable concepts there. However, I think it's being overtaken with hype and I worry that hype will hurt the credibility of the truly valid concepts. When I see engineers with real RF experience making the Open Spectrum movement's claims, I will be much less skeptical.
--zawada
No, the FCC requires you to know code to legally operate on HF because an international treaty requires them to. I believe there is a proposal on the table at WRC-03 to remove that requirement.
--zawada