I don't think they (or at least should) use the frequency to measure the load. You can trust the grid to supply an average frequency as specified (50/60 Hz).
On the contrary, frequency is an excellent indicator of power system health. Proportioning the load to observed frequency is a simple and potenially very effective method of improving power system stability.
Appliances aren't that picky about frequency (with the exception of synchronous motors used for tape machines and turntables) - I've seen motors with 50 Hz on the nameplate run very happily at 60 Hz, this motor was made in the US for use in the US.
The long term stability of the frequency reference is not critical - something on the order of 100 ppm should work nicely (my Timex watch does better than 10 ppm). A simple device that drops the load when the local frequency dips and picks up when the frequency rises would do wonders for improving grid stability as it acts to damp out the oscillations in the power system (think huge weights connected by springs - the frequency sensitive load acts as friction).
Another way to describe what's being done is load shedding on a small scale - back in the mid 1970's the load shedding schedule for California was as folows:
60.00 Hz: Normal operations
59.75 Hz: State water project pumps go off-line
59.50 Hz: Interuptible customers go off-line
59.25 Hz: Rolling Blackouts
This brought back memories of alt.destroy.microsoft from the mid-1990's - reference to their software being Broken As Designed - this is a classic example.
Rather ironic that no-one else has replied to your post - you brought up a very good point.
I've seen a few references stating that the major reason that there hasn't been a major war in Europe is due to it being occupied by the US and Soviets for 45 years (1945 to ca 1990). Strangely enough, the war in Bosnia started just after the occupation was winding down.
Then again, the European theater of WW2 wouldn't have happened if France wasn't so hell bent on punishing Germany for WW1.
Just how many lives were lost in WW1 and WW2? Remember that the US did not want to get involved in either WW1 or WW2 - and probably should have stayed out of WW1.
Assuming that the comment on county imposed power/floor space limits aren't Hemos's idea of a joke: The county may have good reason to limit how much power is being dissapated per square foot - cooling usually means air conditioning and there can be hard limits about how many compressors you can put on a building.
The services database is new to Solaris 10 - previous versions used the usual collection of start-up scripts and inet.conf - besides this is just for the root level stuff. The userland databases are in the users home directory (e.g. ~/.dt).
The other aspect is that the pop-ups have a Windoze look about them - not very convincing when you're running CDE/Motif.
The other area where IBM needs to catch up with Sun is software licensing - I can download Solaris 10 and Studio 11 for free - support does require real money ($120 per socket per year for Sol 10 and somethinh like $900/year for Studio 11). I would be very surprised if IBM's compiler collection wasn't able to run rings around GCC for AIX on Power - having IBM match Sun's licensing could do wonders for the number of people porting to AIX.
What would make things really interesting if Sun carries out its threat to port Solaris to Power (Joerg Schilling is already porting his OpenSolaris distro, SchilliX, to PowerPC). Sun did have a port back in the Sol 2.5.1 days which included the development tools.
I would bet that shoring up AIX is in repsonse to Sun's greater emphasis on Solaris.
I would amend that to say "Sun's greater emphasis on Solaris for x86, specifically AMD64". The Opteron boxes have very nice price/performance.
It will be interesting to watch them duke it out - done right, both may benefit. If IBM really wants to compete, they need to start selling low cost workstations to encourage further development on AIX. This could prove beneficial to the OSS community as well - having to debug porting issues with software is probably much more effective at maintaining the quality of code than having thousands of eyeballs looking over the code.
You know the old Saturday Night Live line, "We're the phone company, we don't have to care"?
That was Ernestine (Lily Tomlin) from Rowan and Martin's Laugh-In not from Saturday Night Pre-Recorded For Your Time Zone. Laugh-in was NBC's reply to the Smothers Brothers Variety Hour (on CBS) - you'll have to buy the DVD collection of the first year of the Muppet show for a reference to ABC's response.
BTW, there were some excellent jabs at TPC (The Phone Company) in the movie "The President's Analyst".
FWIW, most companies want some sort of regulation - telco's and cable companies don't want competitors muscling in on their turf - witness BellSouth's reaction to New Orleans bid to provide municipal wi-fi and the cableco's reaction to Verizon offering TV over VDSL and FIOS.
I believe that you could buy a Data General Nova with 64K of memory (core) for about $15,000 in 1972, but that didn't include a terminal (KSR-33's were the least expensive at that time) or mass storage (paper tape would have been the chepaest solution). A few people did buy such computers for home use. IBM had just released the 8 inch floppy to replace the funky System 3 punch cards (128 characters per card) which were intended to replace the 80 column Hollerith cards - incidentally the floppy had about the same data storage density of a box of hollerith cards.
The chief enabling technologies (as you hinted) for the PC was low cost RAM, low cost mass storage and low cost display.
The Stanley was quite a car in its time - the land speed record was held by a Stanley for a short while (ISTR ~1906). An even better steamer was the Doble from the late twenties - a few of them are still driveable (and not many were made).
What may make more sense is the recent devlopment in high performance Seebeck (sp?) effect devices (aka thermocouples) - probably a lot less mass and the electric power output is probably more useful than the mechanical output of a steam engine.
What BMW is doing is more long the lines of the combined cycle power plants - where the exhaust heat from gas turbines are used to make steam for steam turbines. The Stanley Steamer is more akin to a conventional steam plant.
Curtis-Wright did something similar with the turbo-compound engines, where exhaust turbines were coupled to the crankshaft - got about 20% more power for a given fuel consumption - and allowed the DC-7C and L-1649's to go from New York to London/Paris nonstop.
Microsoft has produced sophisticated hardware before, for example Z80 coprocessor cards for Apple IIs. This let Apple II users run CP/M back in the day.
The Z-80 was designed by Seattle Computer Products (the same folks who originally wrote what is now MS-DOS), not MS. One interesting "feature" was that MS was selling the card with their Basic for $295 at the same time they were charging $295 for Basic running on a run of the mill CP/M system - they added some hooks to the Z-80 card Basic so that it would only run on the Z-80 card.
Not sure if you've heard, MS does not make any mention of the CPU's power requirements - word is that the yield of their version of the Cell is lower than desired - and they are upping the voltage (hence power) to get a larger pile of parts to pick from. It is also possible that the PSU was spec'ed for Cell's that met the spec (and didn't need higher voltage) and is straining from having to drive an overclocked processor.
The focus of TFA is how Google's bucking the trend in the world of mergers and acquisition and venture capital, which has in turn ruffled the feathers of more entrenched high-tech business interests.
Especially when considering that M&A are what makes most of the money for investment bankers. As others have pointed out, TFA did come across as sour grapes from the "financial world".
I will say that Google is way overpriced at the moment, pretty much the same way that Sun was overpriced at its peak.
In a word, no. Binary compatibility just isn't. It happens to work in very select situations, usually on proprietary operating systems that are themselves binary-only, which means the developers of the OS would be dealing with the same headaches you would if they broke binary compatibility.
There are a couple of paths that proprietary OS's have taken to maintain binary compatibility. One approach taken by Apple and Sun is to thoroughly document what will be the stable ABI and maintain backwards compatibility with that ABI. The other is the Microsoft approach which is to have purposely incomplete documentation on the ABI and then struggle to accomondate applications that use the undocumented ABI's. Linux, while having a well documented API, doesn't necessarily guarantee backwards compatibility (then again, neither did the original BSD - original as in from UCB).
I've had a similar experience with dealing with pine and outlook, it is much faster saving a message to a folder with pine even though I'm a terrible typist.
A lot of the young-uns around here would do well to look back at the CLI vs GUI debate that took place around the introduction of the Mac.
Maybe vibrations mess up the flow of electrons in the tubes?
That's not as far fetched as it may sound. The flow of electrons in the tubes are governed by the voltage on the grid wires and the spacing of the grid wires - the latter can be affected by vibrations (microphonics). In addition, most ceramic capacitors show some piezoelectric effect (exceptions are porcelain ceramics).
Tubes do have an advantage over transistors - the interelectrode capacitances are not dependent on operating voltages.
I'm hoping these are not "standard" shipping containers, just something that looks like them.
A bit more to the point, the box needs have the same critical dimensions of a shipping container - specifically the mounting points. The dimenions, BTW, are in feet and inches since the ISO container was a North American invention of the late 50's and intended to be the equivalent of a highway trailer without the chassis (so it could be put on truck, trains and ships).
What about if they use dual-core low-power Opterons (the HE models), which use an amount of power comparable to single-core models ? And hop! suddenly you end up with a 50W per logical CPU.
I seem to recall someone pointing out that a single Niagara server could easily keep up with a quad Opteron box (don't remember if they were single or dual core Opterons). One thing to bear in mind is that the power draw for the memory system is now approaching the CPU's draw - in the case of the Niagara, I suspect a full up memory will draw more than the CPU.
Wondering if this is why Sun was working so hard to get the Niagara out ahead of schedule?
Slashdot Mirror
On the contrary, frequency is an excellent indicator of power system health. Proportioning the load to observed frequency is a simple and potenially very effective method of improving power system stability.
Appliances aren't that picky about frequency (with the exception of synchronous motors used for tape machines and turntables) - I've seen motors with 50 Hz on the nameplate run very happily at 60 Hz, this motor was made in the US for use in the US.
Another way to describe what's being done is load shedding on a small scale - back in the mid 1970's the load shedding schedule for California was as folows:
60.00 Hz: Normal operations
59.75 Hz: State water project pumps go off-line
59.50 Hz: Interuptible customers go off-line
59.25 Hz: Rolling Blackouts
Emulating O. Winston Link shooting J-1's and Y-6b's?
Stuff that turns your skin brown???
IIRC, most films use AgI...
IOW, we're still about the same place as we were in Newton's time. OTOH, that means there are still some open frontiers in science.
Thanks for the Larry Seltzer link.
Does that mean that cdrw will be updated to support the Blu-Ray drives?
I've seen a few references stating that the major reason that there hasn't been a major war in Europe is due to it being occupied by the US and Soviets for 45 years (1945 to ca 1990). Strangely enough, the war in Bosnia started just after the occupation was winding down.
Then again, the European theater of WW2 wouldn't have happened if France wasn't so hell bent on punishing Germany for WW1.
Just how many lives were lost in WW1 and WW2? Remember that the US did not want to get involved in either WW1 or WW2 - and probably should have stayed out of WW1.
Assuming that the comment on county imposed power/floor space limits aren't Hemos's idea of a joke: The county may have good reason to limit how much power is being dissapated per square foot - cooling usually means air conditioning and there can be hard limits about how many compressors you can put on a building.
The other aspect is that the pop-ups have a Windoze look about them - not very convincing when you're running CDE/Motif.
What would make things really interesting if Sun carries out its threat to port Solaris to Power (Joerg Schilling is already porting his OpenSolaris distro, SchilliX, to PowerPC). Sun did have a port back in the Sol 2.5.1 days which included the development tools.
I would amend that to say "Sun's greater emphasis on Solaris for x86, specifically AMD64". The Opteron boxes have very nice price/performance.
It will be interesting to watch them duke it out - done right, both may benefit. If IBM really wants to compete, they need to start selling low cost workstations to encourage further development on AIX. This could prove beneficial to the OSS community as well - having to debug porting issues with software is probably much more effective at maintaining the quality of code than having thousands of eyeballs looking over the code.
That was Ernestine (Lily Tomlin) from Rowan and Martin's Laugh-In not from Saturday Night Pre-Recorded For Your Time Zone. Laugh-in was NBC's reply to the Smothers Brothers Variety Hour (on CBS) - you'll have to buy the DVD collection of the first year of the Muppet show for a reference to ABC's response.
BTW, there were some excellent jabs at TPC (The Phone Company) in the movie "The President's Analyst".
FWIW, most companies want some sort of regulation - telco's and cable companies don't want competitors muscling in on their turf - witness BellSouth's reaction to New Orleans bid to provide municipal wi-fi and the cableco's reaction to Verizon offering TV over VDSL and FIOS.
The chief enabling technologies (as you hinted) for the PC was low cost RAM, low cost mass storage and low cost display.
What may make more sense is the recent devlopment in high performance Seebeck (sp?) effect devices (aka thermocouples) - probably a lot less mass and the electric power output is probably more useful than the mechanical output of a steam engine.
Curtis-Wright did something similar with the turbo-compound engines, where exhaust turbines were coupled to the crankshaft - got about 20% more power for a given fuel consumption - and allowed the DC-7C and L-1649's to go from New York to London/Paris nonstop.
How about four?
The Z-80 was designed by Seattle Computer Products (the same folks who originally wrote what is now MS-DOS), not MS. One interesting "feature" was that MS was selling the card with their Basic for $295 at the same time they were charging $295 for Basic running on a run of the mill CP/M system - they added some hooks to the Z-80 card Basic so that it would only run on the Z-80 card.
Not sure if you've heard, MS does not make any mention of the CPU's power requirements - word is that the yield of their version of the Cell is lower than desired - and they are upping the voltage (hence power) to get a larger pile of parts to pick from. It is also possible that the PSU was spec'ed for Cell's that met the spec (and didn't need higher voltage) and is straining from having to drive an overclocked processor.
Especially when considering that M&A are what makes most of the money for investment bankers. As others have pointed out, TFA did come across as sour grapes from the "financial world".
I will say that Google is way overpriced at the moment, pretty much the same way that Sun was overpriced at its peak.
There are a couple of paths that proprietary OS's have taken to maintain binary compatibility. One approach taken by Apple and Sun is to thoroughly document what will be the stable ABI and maintain backwards compatibility with that ABI. The other is the Microsoft approach which is to have purposely incomplete documentation on the ABI and then struggle to accomondate applications that use the undocumented ABI's. Linux, while having a well documented API, doesn't necessarily guarantee backwards compatibility (then again, neither did the original BSD - original as in from UCB).
The 6 year old from TZ is now 51.
A lot of the young-uns around here would do well to look back at the CLI vs GUI debate that took place around the introduction of the Mac.
That's not as far fetched as it may sound. The flow of electrons in the tubes are governed by the voltage on the grid wires and the spacing of the grid wires - the latter can be affected by vibrations (microphonics). In addition, most ceramic capacitors show some piezoelectric effect (exceptions are porcelain ceramics).
Tubes do have an advantage over transistors - the interelectrode capacitances are not dependent on operating voltages.
A bit more to the point, the box needs have the same critical dimensions of a shipping container - specifically the mounting points. The dimenions, BTW, are in feet and inches since the ISO container was a North American invention of the late 50's and intended to be the equivalent of a highway trailer without the chassis (so it could be put on truck, trains and ships).
Nice posting.
I seem to recall someone pointing out that a single Niagara server could easily keep up with a quad Opteron box (don't remember if they were single or dual core Opterons). One thing to bear in mind is that the power draw for the memory system is now approaching the CPU's draw - in the case of the Niagara, I suspect a full up memory will draw more than the CPU.
Wondering if this is why Sun was working so hard to get the Niagara out ahead of schedule?