2.4 GHz frequency is optimal for warming food in microwave ovens.
Without reading the article - I would guess that "optimality" is more of a good compromise - penetration gets better at lower frequencies (reason for using 916 MHz in the "industrial microwave oven") and being able to "brown" the foods by surface heating. Another reason is smaller nodes in the standing waves inside the oven cavity.
Heck, I remembering reading an article from an early 1970's issue of QST about a resonant cavity made with a garbage can (running at 146 MHz) - and it worked very well at cooking hamburgers between te plates of the tuning capacitor - heating coming from dielectric absorption (which is much more pronounced at 2.4 GHz than 146 MHz).
Oh, and the reason for 2.4 GHz being usable frequency world wide: it happens to be water absorption frequency (which happens to be the technical detail in microwave ovens), hence it being unusable for traditional frequency owners.
BZZZZT!! The water absorption frequency is up around 21 GHz and the liquid absorption line is REALLY broad. I have seen a large industrial "microwave oven" running at 916 MHz - so there is nothing magical about 2.4 GHz.
The reason that 2.4 GHz is widely used is that most countries have agreed on the use for ISM (Industrial, Scientific, Medical). Other users of the frequency must tolerate interference from these sources, which makes it undesirable for licensed services.
If I broadcast on a spectrum being used by another, it can interfere. And by interfering, I can exclude others from using it.
That's the reason for regulation of the radio spectrum. (Score:5 insightful)
A critical point that Shirky and countless others fail's to graps is that the reasons unlicensed spectrum works are:
1) The frequency band in use does not support long distance propagation (i.e. ionspheric reflection) - hence interference is a local problem.
2) The FCC and equivalent regulatory bodies limit the transmit power (and often effective radiated power).
There is a simple method for minimizing interference - use directional antennas with the narrowest possible beamwidth (some of this can be done electronically) for both transmit and receive - and lowest power needed for communication. Unfortunately, most wi-fi users use omnidirectional antennas.
I've also had w-a-y too much experience with interference with poorly designed consumer electronics and similar problems to think that the FCC and related organizations are obsolete. Sure, the state of the art equipment is less susceptible than previous generations of equipment, but I strongly doubt that the typical consumer will anything close to state of the art. To prove my point on the latter - look at all of the problems on the internet due to viruses, worms and the ilk that propagate through Winoze boxes...
Saf[e]ty and reliability are absolutes in industrial control.
The focus of this work is is developing systems for monitoring, not control. The wireless technology is to avoid the cost of running wires, the mesh aspect is to minimize the number of access points.
The technology is no where near established enough for safety critical systems.
I'd agree with you if the wireless mesh was being used for process control.
In this case, the wireless mesh is being used for monitoring the health of the motor not for active process feedback.
Car analogy: When driving a car, your primary feedback is what you see out the windows and that's what your attention is focused on 99% of the time. Every now and then, you take a look at the gauges to make sure the engine is running properly (which is what the info on the mesh provides). If the gauges show something amiss, you may have anywhere from a few seconds (pegged coolant temp or no oil pressure) to several hours (or longer) to deal with the problem.
Which is pretty much the same as being able to stop distribution at will.
This, IMHO, may be the argument that will doom the DMCA. Say someone writes a piece that ticks off MegaCorp - MegaCorp threatens all sorts of legal action until the author signs over the copyright to MegaCorp (M$ vs Sprynet, katie.com, DirectTV, etc for similar situations) - then MegaCorp stops distribution of "their" copyrighted material. No more freedom of speech.
Another way of attacking the DRM issue is to declare that anything protected by DRM is considered "commercial speech".
Effectively, it will drive the production of recording devices to countries where
they are not regulated.
That's basically why Ampex didn't get into the VCR market, even though they did invent the VTR (Video Tape Recorder) back in the mid-50's. Sony had a lot more money to fight the Hollyweird studios.
A simple solution to this problem has been proposed: Any work not available from the copyright owner for a year or more should lose its copyright and become public domain.
I agree in principle, but, IMHO, a year is too short. Try 3 years for software, 5 years to 10 years for music/movies, and 10 to 15 years for books.
Someone at the Disney Corp got a small clue recently - they will sell replacement DVD's for about $6 - anyone who has young kids knows that CD's/DVD's do not have infinite lifetimes _ Disney does make a bit of money from products aimed at young kids.
but after rotating off the runway, he doesn't need to see anything at all.
I beg to differ. Unless the plane is in controlled airspace, there will be aneed to look out for other traffic (actually not a bad idea even in controlled airspace). The 1978 crash in San Diego took place between two aircraft that were both in contact with air traffic control.
That C compiler is (AFAIK) only shipped for building new versions of the kernel.
One of the more prominent entries on the HP-UX FAQ (more relevant to the days when HP-UX was a contender for desktop UNIX) - was why the built-in compiler was brain-dead. The answer is what you said - it is there for building the kernel (as has been that way since HP-UX 8.x, and probably before). It was a pretty nice compiler as long as the source was strict K&R (it had code to detect ANSI C and would issue an error) - then again, you wouldn't want to trust your kernel to a flaky compiler, would you?
I don't actually see us making a change like this away from Sun simply because there are no true replacements for the types of servers we are using from an x86 standpoint. However, as opterons become more and more available in server class systems, then maybe some of the systems will be converted over, but I don't see this happening anytime in the next 3-4 years
I'd also guess that it will be 3-4 years before Opterons catch up with the current high-end Sun Fire machines - and the vendor of choice may likely be Sun.
If L-M was going to replace Sun Blades with "commodity hardware", then it would make more sense to go with Opteron based workstations than Intel based workstations - epsecially if they will be running Linux. I suspect that Sun will be selling more Opteron workstations that Sparc by year's end (though the 90 nm US-IIIi may keep Sparc in the running).
Hey, I thought Cable and Satelite were more or less restriction free on what they can broadcast.
Since satellite still uses over the air transmission, they are theoretically less "immune" to regulation of content than cable. A cable-co can block objectionable material from even getting into your house by filtering out the appropriate channels.
Methinks the broadcasters are now reaping what they've sown - they've been asking for increased regulation in the form of "broadcast flags" and the like - the FCC is now saying "Oh you want more regulation? WE'LL give you more regulation!"
The theme for the November 1984 issue of Unix Review was "User Friendly Systems" - one of the comments was that using nroff with the memo macro package could be a lot more productive than using MacWrite's GUI formatting. Learning to use nroff was a lot harder than learning to use MacWrite.
Similarly, learning to use regular expressions takes a bit of effort, but once learned they make the command line a lot more useful.
Re:Youwant dual mode transportation?
on
By Road and Rail?
·
· Score: 4, Informative
The common American term is "piggybacking" - more officially known as TOFC (trailer on flat car).
The modern implementation of TOFC started in the mid-1950's - special flats cars were being built in the early 1960's (often owned by Trailer-Train) - the earliest implementation dates back to about 1920, didn't take off then because of opposition from the state highway authorities (trucks were avoiding road use fees).
For long hauls - it makes more sense just to use the box (i.e. containers) - as it reduces weight and air resistance. The Espee pioneered double-stacks (i.e. stacking two containers on one car) with articulated car-sets to further reduce tare weight and train length (single stack trains were too long for the sidings).
To answer your question - the onde advantage of this approach over TOFC is that you can have much smaller trains.
Re:Youwant dual mode transportation?
on
By Road and Rail?
·
· Score: 1
Thats what roadrailers are for.
I agree with one exception - it would be nice to have a low cost locomotive for pulling maybe a dozen Roadrailers down some branchlines.
What the parent is alluding to are "field of use" restrictions. More so than many other countries, the U.S. requires that a person registering a trademark provide fairly specific fields in which the trademark is being used or will soon be used.
Tell that to the bastards in intel's legal department - where they have gone after entities way outside of the uProcessor biz for infringing on their trademarks. Microsoft is even worse, but...
Back in the early days of the micro-computer industry, people often referred to hard drives being built with Winchester technology - but I don't recall anyone mentioning that Winchester was a registered tradmark of Olin.
As for the original topic - the Blade Runner sounds like a glorified Hy-Rail truck - and in the US, branch-line track is usually in pretty bad shape.
I'd guess the OP was talking about Dashboard which was modeled after the "dashboard" on VUE (Visual User Environment - which was the basis for CDE). I've found the multiple desktops to be very useful and easy to use (my daughter figured it out at age 2 1/2).
The Photon microGUI on QNX also has had a multiple desktop feature for at least a decade.
If M$ can trademark "Windows", then they probably can patent a lot of prior art. (Sigh)
One of BizWeek's crusades with Sun is "drop Solaris for Linux" claiming that will save a lot of R&D money. BizWeek misses the point in three major areas - one is that there would still be a need for R&D to support the hardware - the second is that Solaris already leverages off of open source code (e.g. Gnome) - and the third is that there is a lot more to open source than Linux (e.g. BSD) - I wonder if Sun is more afraid of OpenBSD than Linux.
I don't remember reading much about OpenOffice.org in BizWeek.
In regards to another posting of yours: 60 bits > 36 bits (clue: what hardware ran SCOPE for an OS)
There's a well-known story about HP buying Microsoft's assurance, in 1995, that MS would soon bring out an enterprise quality NT for servers as well as workstations.
IIRC, one of the HP execs (initials R.P.) that bought into this went on to SGI, then MS and now someplace else. HP had Lotus port 123 and Ami-Pro to HP-UX and had a nice package in the 9000/712. The latter would have made for a nice business desktop - VUE beat the eff'ing pants off of Win 3.x and upgrade from "sam" was much friendlier than any install utility for Windoze.
Sigh.
Sun lucked out in getting the FPS portion of Cray. They also did a decent job with the "Darwin" workstations of the late 90's - but dropped the ball with the Sun Blade 100 - Sun needs to have good hardware at the low end to remain competitive. They stand a chance with the Opteron boxes (and possibly the 90nm US-IIIi) and they do need to keep the AMD64 port of Solaris on schedule.
I did RTFA last night and wasn't that impressed - get a lot more insight reading comp.arch.
If anyone's demise is fed by Itanic, it would be Intel's and HP's.
No argument from me wrt to intel and HP. It will be interesting to see if BizWeek has anything to say about that.
My point was that Windows on the Itanic was thought by many in the mid-90's to be the future of workstations - thus making Sun irrelevant. Sun was the first major OEM to pull out of the Itanic bandwagon - much to intel's disgust.
Intel's other major boo-boo was trying to push single processor performance too far. With processor speeds now being dominated by interconnect delays, the "smart money" is designing smaller cores and putting more on a chip.
The "reason" that was usually given for Sun's demise from the Wintel onslaught was the Itanic -er- Itanium. Funny how we haven't heard much from the Biz rags about that fiasco.
Along those lines, I'd say that Sun has done much better than HewPaq in the Unix system market.
I think the last chance Sun is going to get is going to be Niagra and Rock.
And I would add getting the AMD-64 port of Solaris out the door on schedule. One of their most critical tasks is maintaining developer mindshare for Solaris.
Slashdot Mirror
Without reading the article - I would guess that "optimality" is more of a good compromise - penetration gets better at lower frequencies (reason for using 916 MHz in the "industrial microwave oven") and being able to "brown" the foods by surface heating. Another reason is smaller nodes in the standing waves inside the oven cavity.
Heck, I remembering reading an article from an early 1970's issue of QST about a resonant cavity made with a garbage can (running at 146 MHz) - and it worked very well at cooking hamburgers between te plates of the tuning capacitor - heating coming from dielectric absorption (which is much more pronounced at 2.4 GHz than 146 MHz).
BZZZZT!! The water absorption frequency is up around 21 GHz and the liquid absorption line is REALLY broad. I have seen a large industrial "microwave oven" running at 916 MHz - so there is nothing magical about 2.4 GHz.
The reason that 2.4 GHz is widely used is that most countries have agreed on the use for ISM (Industrial, Scientific, Medical). Other users of the frequency must tolerate interference from these sources, which makes it undesirable for licensed services.
That's the reason for regulation of the radio spectrum. (Score:5 insightful)
A critical point that Shirky and countless others fail's to graps is that the reasons unlicensed spectrum works are:
1) The frequency band in use does not support long distance propagation (i.e. ionspheric reflection) - hence interference is a local problem.
2) The FCC and equivalent regulatory bodies limit the transmit power (and often effective radiated power).
There is a simple method for minimizing interference - use directional antennas with the narrowest possible beamwidth (some of this can be done electronically) for both transmit and receive - and lowest power needed for communication. Unfortunately, most wi-fi users use omnidirectional antennas.
I've also had w-a-y too much experience with interference with poorly designed consumer electronics and similar problems to think that the FCC and related organizations are obsolete. Sure, the state of the art equipment is less susceptible than previous generations of equipment, but I strongly doubt that the typical consumer will anything close to state of the art. To prove my point on the latter - look at all of the problems on the internet due to viruses, worms and the ilk that propagate through Winoze boxes...
The focus of this work is is developing systems for monitoring, not control. The wireless technology is to avoid the cost of running wires, the mesh aspect is to minimize the number of access points.
The technology is no where near established enough for safety critical systems.
In this case, the wireless mesh is being used for monitoring the health of the motor not for active process feedback.
Car analogy: When driving a car, your primary feedback is what you see out the windows and that's what your attention is focused on 99% of the time. Every now and then, you take a look at the gauges to make sure the engine is running properly (which is what the info on the mesh provides). If the gauges show something amiss, you may have anywhere from a few seconds (pegged coolant temp or no oil pressure) to several hours (or longer) to deal with the problem.
Which is pretty much the same as being able to stop distribution at will.
This, IMHO, may be the argument that will doom the DMCA. Say someone writes a piece that ticks off MegaCorp - MegaCorp threatens all sorts of legal action until the author signs over the copyright to MegaCorp (M$ vs Sprynet, katie.com, DirectTV, etc for similar situations) - then MegaCorp stops distribution of "their" copyrighted material. No more freedom of speech.
Another way of attacking the DRM issue is to declare that anything protected by DRM is considered "commercial speech".
That's basically why Ampex didn't get into the VCR market, even though they did invent the VTR (Video Tape Recorder) back in the mid-50's. Sony had a lot more money to fight the Hollyweird studios.
I agree in principle, but, IMHO, a year is too short. Try 3 years for software, 5 years to 10 years for music/movies, and 10 to 15 years for books.
Someone at the Disney Corp got a small clue recently - they will sell replacement DVD's for about $6 - anyone who has young kids knows that CD's/DVD's do not have infinite lifetimes _ Disney does make a bit of money from products aimed at young kids.
I beg to differ. Unless the plane is in controlled airspace, there will be aneed to look out for other traffic (actually not a bad idea even in controlled airspace). The 1978 crash in San Diego took place between two aircraft that were both in contact with air traffic control.
One way to increase security would be using a range of UDP ports, encrypting a timestamp and perhaps serial number to prevent replay attacks.
One of the more prominent entries on the HP-UX FAQ (more relevant to the days when HP-UX was a contender for desktop UNIX) - was why the built-in compiler was brain-dead. The answer is what you said - it is there for building the kernel (as has been that way since HP-UX 8.x, and probably before). It was a pretty nice compiler as long as the source was strict K&R (it had code to detect ANSI C and would issue an error) - then again, you wouldn't want to trust your kernel to a flaky compiler, would you?
QNX is still being used - the latest version now hs gcc support - versus that Watcom compiler for earlier versions (4.25 and earlier).
I'd also guess that it will be 3-4 years before Opterons catch up with the current high-end Sun Fire machines - and the vendor of choice may likely be Sun.
If L-M was going to replace Sun Blades with "commodity hardware", then it would make more sense to go with Opteron based workstations than Intel based workstations - epsecially if they will be running Linux. I suspect that Sun will be selling more Opteron workstations that Sparc by year's end (though the 90 nm US-IIIi may keep Sparc in the running).
Since satellite still uses over the air transmission, they are theoretically less "immune" to regulation of content than cable. A cable-co can block objectionable material from even getting into your house by filtering out the appropriate channels.
Methinks the broadcasters are now reaping what they've sown - they've been asking for increased regulation in the form of "broadcast flags" and the like - the FCC is now saying "Oh you want more regulation? WE'LL give you more regulation!"
I couldn't agree more.
The theme for the November 1984 issue of Unix Review was "User Friendly Systems" - one of the comments was that using nroff with the memo macro package could be a lot more productive than using MacWrite's GUI formatting. Learning to use nroff was a lot harder than learning to use MacWrite.
Similarly, learning to use regular expressions takes a bit of effort, but once learned they make the command line a lot more useful.
The modern implementation of TOFC started in the mid-1950's - special flats cars were being built in the early 1960's (often owned by Trailer-Train) - the earliest implementation dates back to about 1920, didn't take off then because of opposition from the state highway authorities (trucks were avoiding road use fees).
For long hauls - it makes more sense just to use the box (i.e. containers) - as it reduces weight and air resistance. The Espee pioneered double-stacks (i.e. stacking two containers on one car) with articulated car-sets to further reduce tare weight and train length (single stack trains were too long for the sidings).
To answer your question - the onde advantage of this approach over TOFC is that you can have much smaller trains.
I agree with one exception - it would be nice to have a low cost locomotive for pulling maybe a dozen Roadrailers down some branchlines.
Tell that to the bastards in intel's legal department - where they have gone after entities way outside of the uProcessor biz for infringing on their trademarks. Microsoft is even worse, but...
Back in the early days of the micro-computer industry, people often referred to hard drives being built with Winchester technology - but I don't recall anyone mentioning that Winchester was a registered tradmark of Olin.
As for the original topic - the Blade Runner sounds like a glorified Hy-Rail truck - and in the US, branch-line track is usually in pretty bad shape.
IIRC, sendmail predates QDOS!
The Photon microGUI on QNX also has had a multiple desktop feature for at least a decade.
If M$ can trademark "Windows", then they probably can patent a lot of prior art. (Sigh)
I don't remember reading much about OpenOffice.org in BizWeek.
In regards to another posting of yours: 60 bits > 36 bits (clue: what hardware ran SCOPE for an OS)
IIRC, one of the HP execs (initials R.P.) that bought into this went on to SGI, then MS and now someplace else. HP had Lotus port 123 and Ami-Pro to HP-UX and had a nice package in the 9000/712. The latter would have made for a nice business desktop - VUE beat the eff'ing pants off of Win 3.x and upgrade from "sam" was much friendlier than any install utility for Windoze.
Sigh.
Sun lucked out in getting the FPS portion of Cray. They also did a decent job with the "Darwin" workstations of the late 90's - but dropped the ball with the Sun Blade 100 - Sun needs to have good hardware at the low end to remain competitive. They stand a chance with the Opteron boxes (and possibly the 90nm US-IIIi) and they do need to keep the AMD64 port of Solaris on schedule.
I did RTFA last night and wasn't that impressed - get a lot more insight reading comp.arch.
No argument from me wrt to intel and HP. It will be interesting to see if BizWeek has anything to say about that.
My point was that Windows on the Itanic was thought by many in the mid-90's to be the future of workstations - thus making Sun irrelevant. Sun was the first major OEM to pull out of the Itanic bandwagon - much to intel's disgust.
Intel's other major boo-boo was trying to push single processor performance too far. With processor speeds now being dominated by interconnect delays, the "smart money" is designing smaller cores and putting more on a chip.
Along those lines, I'd say that Sun has done much better than HewPaq in the Unix system market.
And I would add getting the AMD-64 port of Solaris out the door on schedule. One of their most critical tasks is maintaining developer mindshare for Solaris.