This reduces power losses due to multiple AC/DC conversions,
It is the DC to AC conversion (as in UPS) followed by the AC to DC conversion that gets me. Would make more sense having a large AC/DC power supply feeding a DC bus (with either batteries or UltraCaps) going to DC/DC converters at each board. It is a bit easier getting high efficiency from a moderate input voltage DC/DC converter than a typical AC/DC switcher running off the lines. It is also simplifies having redundant poer supplies.
And many are are content to keep on using Office 97 - just witness the ads MS is running comparing Office 97 users to dinosaurs.
MS seems to be just slightly aware of customers needs in data retention - especially for governments (So what's the exact wording of the law I'm supposed to be violating and what do you mean that the law is in a now unreadable version of MS-Word??). Small case in point, the City of Encinitas used to use MS-Word to publish the Municipal Code on their website - now it is posted as PDF's.
The biggest problems are keeping it together, and keeping it protected from harm, like accidently hitting it in a plane, or lightning strikes. It could become a terrorist target.
What just about everybody forgets about the spece elevator is that every orbit lower than geosynchronous will eventually intersect the elevator (assuming the elevator is anchored on the equator). A particle too small to track from earth can still have quite an impact.
One possible solution would be a much better tracking system combined with some method for deflecting/destroying objects that come too close.
I still have a copy of "Assembly Language Programming," by Ralph Grishman (IIRC the author correctly) and it makes no mention of this. Was it a priv. instruction?
I just checked my copy (and you got Grishman's name right) and it said that accessing ECS would not be discussed (page 15 in my copy).
I do remember seeing A0:X0 being mentioned on comp.sys.cdc in regards to addressing ECS. It is also possible that mention of ECS may be in the CDC 6000 series manual I have lying around - somewhere. IIRC, A0 pointed to the start of the block in the main memory for ECS I/O and X0 was packed with the ECS address and number of words to move (none of this byte crap) - but I may have remembered incorrectly.
B0, of course, was set to o000000.
The memories of using the "RUN" compiler, "compass" which stood for COMPrehensive ASSembler or was it COMPlete pain in the ASS?, LGO and CLDR.Another memory was diving into the recycle bin and seeing the listing for SPICE.
Memories of using a CRT based terminal (Hazeltine) with a data cassette local storage (with some thoughts of using those for mass storage for the SS-90 just below). Seeing the side of the terminal open revealing the banks of core.
For my first quarter at the big U, we had access to the 6400 "B" machine running KRONOS. The machine got sold - but the CS department was able to get hold of a PDP-11/44 for interactive use - and a year laer started playing with a tape from Bell Labs. By that time, my interest was in circuits, so never got to know some of the people playing the PDP-11's. IIRC, the Hazeltines were one of the terminal types that got connected to the PDP-11's (the 11/70 was in the same room as the Hazeltine's), prompting the CS guys to come up with a way to support many terminal types (too bad their idea never took hold;-^).
More memories were picking up a copy of Byte when starting grad school and wondering what it would be like owning my own computer - going across the bay to see the first West Coast Computer Faire (and finding out that one of my high school classmates attended as well - ended up working for Apple within a year - I probably ran into one of the Steve's - got one of the original Apple II flyers).
The San Francisco computer shows of the early 80s. Those were fun shows.
Still have fond memories of the first West Coast Computer Faire of 1977 - back in the days when small systems were pretty much the realm of the hobbyist.
I still remember that you save data to core using A6 and A7, and retrieve data from core via A1 through A5. A0 was a scratch register. Ah, mainframe assembly.
Hmmm, not quite - A0:X0 were used for transferring data to/from ECS (extended core storage).
I have fond memories of learning machine/assembly for the CDC6000 series - a very clean architecture. There are some people working on a CDC6600 emulator and ISTR hearing about an FPGA implementation.
Also have memories of being enamored by one of the CDC6400 operators at the big U. (A couple of years later, one of my dormates was complaining about some stuff from Bell Labs - "an abomination in the eyes of the lord", "If Bell Labs hadn't invented the transistor, the phone company would still be using vaccuum tubes" - a couple of years after that, one of the guys on campus started sending out tapes with pretty much the same license used for distributing SPICE)
My first exposure to actual computer hardware was at an "open house" at Pt. Mugu - a bit over 5 1/2 years later, I was running my first programs on a CDC1700 (would like to get some documentation on that beast) and wondering what it would be like to have something like that for my own (there were a few weird birds using PDP's and Nova's for home use).
Modern air conditioners have energy efficiencies approaching 400%.
Ummmm, I believe the term you want to use is "coefficient of performance" - which is how many watts of heat are transferred per watt of electrical power used. Also called an energy efficiency ratio.
Having said that, your point about the relative efficiencies of mechanical refrigeration units vs Peltier effect devices is correct. I have a ~18 cu ft fridge in my garage that uses less energy than a 1 cu ft Peltier cooler. Another point, the main focus for the development of Gadolinium refrigeration was to replace Peltier effect devices for small scale refrigeration needs.
The mailbox icon on the dashboard does change when incoming mail s present. The calendar program can be configured to pop-up reminders before an appointment. At one time, CDE was even available for Linux as a commercial product...
The other option would be using xbiff (not to be confused with xbill) and setting up cron to e-mail yerself a reminder.
It doesn't do high end nearly as well as Aix, it doesn't do low end as well as Linux. And most would agree Solaris is better in the middle.
HP-UX started to fall behind Solaris in the mid to late 90's, HP was starting to waffle about the future of HP-UX, while Sun was firmly committed to Solaris. In the early 90's, HP-UX seemed to be a bit more solid than Solaris.
HP also did a good job of pissing off their customers by dumping MPE.
Everything (except capacitors) is a resistive or inductive load.
Over simplified.
You're forgetting about linear vs non-linear loads (think capacitor input power supplies for the latter). I would guess the current waveform in a modern home will look pretty ugly.
The real way to do power monitoring is to sample the current at 4kSamples/sec as well as sample at least one of the voltage wavefforms. Data processing would be taking the average of the instaneous power and recording that average every few seconds.
BTW, watt-hour meters do measure power and not simply current.
For the story about the third bomb - see Richard Rhodes "Making of the Atomic Bomb" - Rhodes specifically said that the third bomb was being shipped by air (this was after the disaster with the Indianopolis - another reason was to get the third bomb there ASAP).
For the story about Truman, see the "New Dealers' War" by Thomas Fleming. FDR had kept Truman in the dark about many things (including the A-bomb work).
THere was an article in "Trains" magazine a couple of years ago about a fellow riding with Truman in 1961 - Truman had stated that they were estimating 95% casualties for the initial assaults on the Japanese homeland - the fellow writing the article would have been on one of the initial assaults.
The one reason that I have sympathy for your argument is that I feel that an entire nation is responsible for its actions
That would imply that the entire nation of Japan was responsible for the "Rape of Nanking" thus justifying killing of Japanese civillians. More civilians were killed in Nanking than in Hiroshima and Nagasaki combined.
consider this as well: we killed far more in March over Tokyo with the firebombings than we did at Hiroshima and Nagasaki.
A couple of factoids:
Truman first found out about the extent of firebombing in June 1945 - and ordered that it be stopped immediatefly. Truman, unlike Roosevelt, did see combat in WW1.
A third bomb was being shipped on August 14th - the plane carrying it was told to stop in northern California.
I haven't met anyone who served in the Pacific theater during WW2 who wasn't grateful that the bombs were used.
Assigning Internet authority to the UN would not only benefit in the short term, but would have gigantic benefits in long term global communication. The benefits from a continual standardized network is beneficial beyond our recognition.
Question for you: Why is "the internet" running TCP/IP (+ UDP/IP, etc.) and not OSI?
Simple reason, TCP/IP was created by technologists who wanted a working system, rather than a bureaucracy. The general rule with IETF was that a working implementation was needed before a standard would be accepted.
Reminds me of a joke about OSI: OSI is to networking as Bary Manilow is to music - something that is supposed to please everyone, but manages to repulsive to everyone.
Apple has said that developers should not depend on the BIOS system being used in a release model. Apple likely threw these dev boxes together quickly and inexpensively, so they haven't had time to design all the extra stuff a final version would have, like Open Firmware.
Open Firmware on x86???? Coool!
Maybe Sun will take the cue and start shipping some of their Opteron boxes with Open Firmware - they're being shipped with PeeCee BIOS - in part to allow Windoze to boot.
As you see, we already did. The fact that our networks are connected to your networks doesn't mean that you built our networks or that they belong to you.
The issue is not the physical network but the protocols. Europe is free to go back to the OSI protocols - and the history of OSI may be one of the reasons that the DOC decided to keep control of the root servers - not to mention the fiasco that the ITU has become.
I was referring to Netscape 7, Solaris and Sparc as being unsupported by Google for Google Maps - but it seems to work pretty well. My point being that Google was being very (unnecessarily?) conservative with their "supported" browser/OS/platform list.
Google Maps just plain doesn't work for me in Firefox
Hmmm. Works for me on Netscape 7.0 on Solaris/Sparc (unsupported browser version on an unsupported OS on an unsupported architecture). I did have to click the "load Google maps anyway" button.
Consider the CF6-50 (old 747-200 engine). It's fuel consumption at cruise (35,000 feet and mach 0.8-ish) is about 0.62 pounds of fuel per pound of thrust per hour. The Concorde engine's consumption at cruise (53,000 feet at mach 2.0) is 1.19 pounds of fuel per pound of thrust per hour. Now, consider that during supersonic cruise, your drag skyrockets by a factor of 2-3. So not only are you using 2x as much fuel for a given level of thrust, but your thrust needs to be twice as high to keep the plane flying!
If the L/D remains the same, then using twice as much fuel per hour to go twice as fast results in the same fuel consumption per mile or kilometer. You are correct in that the L/D is 2-3X lower for a supersonic aircraft.
Engine technology has advanced a bit more than you suggested, especially for supersonic aircraft. Chief example is the F-22 that can fly supersonic without afterburner.
I'm wondering if going to Mach 2.4 is a good idea compared to Mach 1.6. At M1.6, it is possible to get almost isentropic conversion from supersonic flow to subsonic flow with multiple oblique shock waves - which would do wonders for engine performance - now if there was a way of improving L/D.
Another example of gas turbine progress - an early 1950's 4500HP GE gas turbine had a thermal efficiency of ~18%, the latest H series turbines from GE have a 46% thermal efficiency.
I've wondered a little, in the NMR case, if it would be possible to induce nuclear fission or at least some kind of nuclear decay through heavy RF excitation of nuclei.
No.
At least not with static magnetic fields realizable on earth - you might be able to get some interesting stuff happening with magnetar level fields.
OTOH, there is some interesting stuff that happens with nuclear quadrupole moments and gamma ray emissions. The gamma ray emissions show some directionality when the thermal energy drops below the quadrupole resonance energy. This effect is used to calibrate microkelvin thermometers.
Slashdot Mirror
It is the DC to AC conversion (as in UPS) followed by the AC to DC conversion that gets me. Would make more sense having a large AC/DC power supply feeding a DC bus (with either batteries or UltraCaps) going to DC/DC converters at each board. It is a bit easier getting high efficiency from a moderate input voltage DC/DC converter than a typical AC/DC switcher running off the lines. It is also simplifies having redundant poer supplies.
And many are are content to keep on using Office 97 - just witness the ads MS is running comparing Office 97 users to dinosaurs.
MS seems to be just slightly aware of customers needs in data retention - especially for governments (So what's the exact wording of the law I'm supposed to be violating and what do you mean that the law is in a now unreadable version of MS-Word??). Small case in point, the City of Encinitas used to use MS-Word to publish the Municipal Code on their website - now it is posted as PDF's.
What just about everybody forgets about the spece elevator is that every orbit lower than geosynchronous will eventually intersect the elevator (assuming the elevator is anchored on the equator). A particle too small to track from earth can still have quite an impact.
One possible solution would be a much better tracking system combined with some method for deflecting/destroying objects that come too close.
Don't forget that Linux (and several BSD's) will boot up on Sparc boxes as well.
The last PCC, SF Muni 1040, left the factory in 1952.
For those scratching their heads, PCC (President's Conference Commission) was used as a name for the streetcars built in the US from 1936 to 1952.
I just checked my copy (and you got Grishman's name right) and it said that accessing ECS would not be discussed (page 15 in my copy).
I do remember seeing A0:X0 being mentioned on comp.sys.cdc in regards to addressing ECS. It is also possible that mention of ECS may be in the CDC 6000 series manual I have lying around - somewhere. IIRC, A0 pointed to the start of the block in the main memory for ECS I/O and X0 was packed with the ECS address and number of words to move (none of this byte crap) - but I may have remembered incorrectly.
B0, of course, was set to o000000.
The memories of using the "RUN" compiler, "compass" which stood for COMPrehensive ASSembler or was it COMPlete pain in the ASS?, LGO and CLDR.Another memory was diving into the recycle bin and seeing the listing for SPICE.
Memories of using a CRT based terminal (Hazeltine) with a data cassette local storage (with some thoughts of using those for mass storage for the SS-90 just below). Seeing the side of the terminal open revealing the banks of core.
For my first quarter at the big U, we had access to the 6400 "B" machine running KRONOS. The machine got sold - but the CS department was able to get hold of a PDP-11/44 for interactive use - and a year laer started playing with a tape from Bell Labs. By that time, my interest was in circuits, so never got to know some of the people playing the PDP-11's. IIRC, the Hazeltines were one of the terminal types that got connected to the PDP-11's (the 11/70 was in the same room as the Hazeltine's), prompting the CS guys to come up with a way to support many terminal types (too bad their idea never took hold ;-^).
More memories were picking up a copy of Byte when starting grad school and wondering what it would be like owning my own computer - going across the bay to see the first West Coast Computer Faire (and finding out that one of my high school classmates attended as well - ended up working for Apple within a year - I probably ran into one of the Steve's - got one of the original Apple II flyers).
Still have fond memories of the first West Coast Computer Faire of 1977 - back in the days when small systems were pretty much the realm of the hobbyist.
Hmmm, not quite - A0:X0 were used for transferring data to/from ECS (extended core storage).
I have fond memories of learning machine/assembly for the CDC6000 series - a very clean architecture. There are some people working on a CDC6600 emulator and ISTR hearing about an FPGA implementation.
Also have memories of being enamored by one of the CDC6400 operators at the big U. (A couple of years later, one of my dormates was complaining about some stuff from Bell Labs - "an abomination in the eyes of the lord", "If Bell Labs hadn't invented the transistor, the phone company would still be using vaccuum tubes" - a couple of years after that, one of the guys on campus started sending out tapes with pretty much the same license used for distributing SPICE)
My first exposure to actual computer hardware was at an "open house" at Pt. Mugu - a bit over 5 1/2 years later, I was running my first programs on a CDC1700 (would like to get some documentation on that beast) and wondering what it would be like to have something like that for my own (there were a few weird birds using PDP's and Nova's for home use).
Sure glad I went through LV early this month - only hit 108 then.
Ummmm, I believe the term you want to use is "coefficient of performance" - which is how many watts of heat are transferred per watt of electrical power used. Also called an energy efficiency ratio.
Having said that, your point about the relative efficiencies of mechanical refrigeration units vs Peltier effect devices is correct. I have a ~18 cu ft fridge in my garage that uses less energy than a 1 cu ft Peltier cooler. Another point, the main focus for the development of Gadolinium refrigeration was to replace Peltier effect devices for small scale refrigeration needs.
The other option would be using xbiff (not to be confused with xbill) and setting up cron to e-mail yerself a reminder.
HP-UX started to fall behind Solaris in the mid to late 90's, HP was starting to waffle about the future of HP-UX, while Sun was firmly committed to Solaris. In the early 90's, HP-UX seemed to be a bit more solid than Solaris.
HP also did a good job of pissing off their customers by dumping MPE.
Over simplified.
You're forgetting about linear vs non-linear loads (think capacitor input power supplies for the latter). I would guess the current waveform in a modern home will look pretty ugly.
The real way to do power monitoring is to sample the current at 4kSamples/sec as well as sample at least one of the voltage wavefforms. Data processing would be taking the average of the instaneous power and recording that average every few seconds.
BTW, watt-hour meters do measure power and not simply current.
For the story about Truman, see the "New Dealers' War" by Thomas Fleming. FDR had kept Truman in the dark about many things (including the A-bomb work).
THere was an article in "Trains" magazine a couple of years ago about a fellow riding with Truman in 1961 - Truman had stated that they were estimating 95% casualties for the initial assaults on the Japanese homeland - the fellow writing the article would have been on one of the initial assaults.
That would imply that the entire nation of Japan was responsible for the "Rape of Nanking" thus justifying killing of Japanese civillians. More civilians were killed in Nanking than in Hiroshima and Nagasaki combined.
A couple of factoids:
Truman first found out about the extent of firebombing in June 1945 - and ordered that it be stopped immediatefly. Truman, unlike Roosevelt, did see combat in WW1.
A third bomb was being shipped on August 14th - the plane carrying it was told to stop in northern California.
I haven't met anyone who served in the Pacific theater during WW2 who wasn't grateful that the bombs were used.
Question for you: Why is "the internet" running TCP/IP (+ UDP/IP, etc.) and not OSI?
Simple reason, TCP/IP was created by technologists who wanted a working system, rather than a bureaucracy. The general rule with IETF was that a working implementation was needed before a standard would be accepted.
Reminds me of a joke about OSI: OSI is to networking as Bary Manilow is to music - something that is supposed to please everyone, but manages to repulsive to everyone.
Ummm, an 8 inch DSDD disk could hold 1.25MB, while the best with 5 1/4" was 1.2MB. OTOH, by 1987, 3.5" floopies could hold 1.44MB.
Open Firmware on x86???? Coool!
Maybe Sun will take the cue and start shipping some of their Opteron boxes with Open Firmware - they're being shipped with PeeCee BIOS - in part to allow Windoze to boot.
IIRC, HTML is a subset of SGML, agian IIRC developed by some IBM'ers.
My nearly 7 year old U10 is running 5.9.
Hmmm. Works for me on Netscape 7.0 on Solaris/Sparc (unsupported browser version on an unsupported OS on an unsupported architecture). I did have to click the "load Google maps anyway" button.
If the L/D remains the same, then using twice as much fuel per hour to go twice as fast results in the same fuel consumption per mile or kilometer. You are correct in that the L/D is 2-3X lower for a supersonic aircraft.
Engine technology has advanced a bit more than you suggested, especially for supersonic aircraft. Chief example is the F-22 that can fly supersonic without afterburner.
I'm wondering if going to Mach 2.4 is a good idea compared to Mach 1.6. At M1.6, it is possible to get almost isentropic conversion from supersonic flow to subsonic flow with multiple oblique shock waves - which would do wonders for engine performance - now if there was a way of improving L/D.
Another example of gas turbine progress - an early 1950's 4500HP GE gas turbine had a thermal efficiency of ~18%, the latest H series turbines from GE have a 46% thermal efficiency.
At least not with static magnetic fields realizable on earth - you might be able to get some interesting stuff happening with magnetar level fields.
OTOH, there is some interesting stuff that happens with nuclear quadrupole moments and gamma ray emissions. The gamma ray emissions show some directionality when the thermal energy drops below the quadrupole resonance energy. This effect is used to calibrate microkelvin thermometers.