.. and everytime an electron hits a defect in a material, a phonon is produced (well, not always, but often) and thus heat is generated. The point is that for any (sanely designed) optical system, the heat generation will be much lower than for a comparable electrical system.
Heat generation from optical components will not be an issue for the foreseeable future.
LVM can do snapshotting. Check the "lvcreate -s" option. If you're looking for something more backup-like, you might want to take a look at Mike Rubel's rsync-snapshots page. I'm currently using a package called dirvish to do rsync snapshots to my backup server. Works like a charm.:)
Yes, that's true. I think it is quite poorly known that windows supports IPP, though. On usenet you see lots of folks trying to get windows -> samba -> cups printing to work, and frequently noone seems to know that you could just do windows -> cups directly...:(
No doubt you can explain the flaw in their benchmarking process that gave them a factor of 2 speed over sockets.
Factor this or that, but I think it's a clear flaw that he has compared a full-featured production kernel (linux 2.4.20) with his own minimal kernel. It's comparing apples to oranges, IMHO.
Well since you asked..:), I find it somewhat strange that these two topics are bundled together like that. The adiabatic approximation is really QM 101-level material, while understanding Berry phase is IMHO quite difficult. And note, it's "Berry phase", not "Berry's phase". The Berry phase approach is probably most famous for it's central part in the modern quantum mechanical theory of polarization. In case you're interested, a review of the topic can be found in e.g.
Resta: "Macroscopic Polarization in Crystalline Dielectrics: The Geometric Phase Approach" Rev. Mod. Phys. 66, 899 (1994)
I'm waiting for the day that Microsoft Windows GUI will be fully raytrace/radiosity/photon map rendered.
Huh, why settle for petty approximations like that. Just to give an incentive to MS, I'm not happy until the light-matter interaction between the light source and the cursor is given a correct quantum mechanical treatment! Of course, with todays computers that means something on the order of a CPU-year for every nanosecond time evolution in a system with 10-100 atoms. But who cares? You know, do something right or don't do it at all!;-)
Um, could you clarify what you mean by "active" vs. "dead" objects?
I meant that with hibernate, if I have a persistent class, say
class Foo { Bar B; // setB and getB methods }
I don't need to explicitly manipulate B through the hibernate API, it works transparently. Eg. if I have loaded an object of the class Foo (say foo), I can access B directly
foo.getB();
Depending on how you configure it, B is loaded at the same time as foo, or it is backed by a proxy and loaded only when it is accessed for the first time.
Of course, being backed by a RDBMS, fancy shit like runtime addition of properties to a class etc. possible with e.g. metaclass hacking in python or MOP in LISP aren't possible. Anyway, if your requirements are on this level, I think Java is the wrong choice for implementation language (aspectj etc. notwithstanding).
Here's a port of prevayler to python: PyPerSyst. The basic concept is rather simple, so I don't think it would be very hard to implement in language X, in case you're interrested.
However, when you have more active objects- which generally arise when you are actually doing good OO design and programming- the interface falls apart some. What if I've got my data objects pointing to other complex objects, rather than just elemental/translatable types like strings, numbers, and dates? What if the state of object A depends directly on what is going on in object B?
In an OORDBMS translation system, the objects are essentially dead while in storage. Object A can point to Object B - another row in a different database which holds data from a different class- but it dead.
In my experience, hibernate solves these kinds of problems quite neatly, with support for collections, dependent objects and whatnot. I'm sure other OR mappers worth anything do the same.
Uh oh, guess how myrinet, quadrics and scali achieve their indeed impressively low latency? By having special user-space MPI libraries that access the hardware directly, without the kernel. And of course, they dont use the IP protocol, but some proprietary protocol designed specifically for cluster use (as simple as possible e.g. no routing)
So, unfortunately the technology used for cluster interconnects is totally non-general purpose. Actually it's more or less useless unless you have a MPI application.
I'm a bit sick of reactionary fear of technology. I work for a major university and deal with "outcry" to many of our "potentially dangerous" research projects.
No shit. There was recently a case of an experiment to study vortexes in liquid He or something like that. Anyway, the equations that describe the flow of this fluid is very similar to the equations that describe black holes.. Yeah, you guessed it; some enviro-wackos tried to prohibit the experiment because they were afraid that a black hole would be created which would swallow the earth.
*sigh* indeed.
I hate to tell the reactionaries this, but the people capable of, say, bioengineering plants to extract toxins from the soil, are also the most competent ones for putting in safeguards and policing themselves.
It seems to me that these days the green movement is more about micromanaging peoples lives than about truly caring about the enviroment. I find it very sad.
The world is actually already full of self-replicating nanoscale bots, and at this very moment your very own precious body contains billions of them! Scary, huh? Better outlaw them until they do something dangerous!
They are called "bacteria". They have been around long before us, and they will be here long after the last human has died.
Need to re-read your facts, violent crime has increased as gun ownership has decreased. Conversely in areas of the US where *legal* gun ownership increases, violent crime decreases.
So, where in my post did I say anything that would contradict the above statement? I just pointed out the difference in gun homicide rates.
I didn't say that creating strict gun control laws would magically solve the problem of high gun homicide in the US. In fact, I believe it might not, because an important part of gun control is to make it much harder for criminals to aquire guns. With hundreds of millions of unlicenced guns already circulating in the US, I suspect there might actually be an increase in gun crime because the criminals (who would have no trouble getting guns anyway) would take advantage of the less well armed populace.
I guess that was sort of your point anyway, huh?
You also forgot to mention the appalling reduction of rights in the Britain scenario.
Reduction of what rights? Personally, I'm more afraid of gun nuts shooting at everything that moves during the night because they're afraid of the dark, than of my government turning against me.
And besides, if the government turns against me, what good would it do if I have a handgun and poppas shotgun when they're coming at me with tanks, choppers, GPS-guided bombs and whatnot?
Regardless of the accuracy of your numbers, you actually want to suggest that we trade trade liberty for perceived safety ( and governmental control )?
Well, since I don't live in the USA and have no plans of moving there either, I don't give a flying fuck. If your vision of liberty is that everybody walks around armed to the teeth constantly afraid of each other and the big bad government, so be it. If you're happy with it, good for you.
You sound like a socialist to me...
Because I quoted some figures on gun homicide rates in Britain and the USA? So, everybody that says something that doesn't shine a bright light on the USA is a commie swine? Your vision of liberty doesn't include freedom of thought?
Or, perhaps your extraordinary vision foresaw that I truly am a commie, evil down to the bottom of my rotten heart. BOOH! HISS! Better start digging that foxhole in your backyard before those evil commie tank armies roll you over.;-)
Everyone gets a couple of solar panels on the roof for the high-pressure system, and a small turbine for the low-pressure.
Uh, remember economics of scale. A big wind turbine will produce electricity cheaper than many small ones. For solar cell though, your argument is correct. As solar cells get cheaper, the day will certainly come where they will be placed on most roofs. Currently solar power is outrageously expensive though, so it will take a long time.
I have such a thing in my appartment. And you know what? I never use the low-flush button. Why? Because if I do, the toilet fouls up so fast that you have to clean it twice a week. Ugh. Another factor being that I don't pay a separate water bill, and where I live the water supply is abundant. The water company even has to run water through the mains pipes sometimes to avoid impuritities sticking to the walls of the pipes.
Anyway, as the previous poster said, a more useful system than these low-flush toilets would be to utilize gray water.
Slashdot Mirror
Can opterons come in 106 processor combinations?
Yes. Will 10000 do?
No, it's not available yet. But the point, I think, is that it is feasible to put lots of opterons in one box.
.. and everytime an electron hits a defect in a material, a phonon is produced (well, not always, but often) and thus heat is generated. The point is that for any (sanely designed) optical system, the heat generation will be much lower than for a comparable electrical system.
Heat generation from optical components will not be an issue for the foreseeable future.
The only reason I'd go with an nVidia card is for their Linux support.
Hrm. I recently switched from a geforce 2mx to a radeon 9200. I chose the radeon instead of a geforce in part because of the Linux support.
The point here being that open source drivers are available for the radeon, while for nvidia you have to use their proprietary binary only driver.
I recently bought a radeon 9200, and I'm quite happy with it. No fan, open source drivers for Linux, good picture quality and it was quite cheap too.
It's about 3x faster than my old geforce 2mx (1980 vs. 650 FPS in glxgears, if that measures anything significant).
LVM can do snapshotting. Check the "lvcreate -s" option. If you're looking for something more backup-like, you might want to take a look at Mike Rubel's rsync-snapshots page. I'm currently using a package called dirvish to do rsync snapshots to my backup server. Works like a charm. :)
Yes, that's true. I think it is quite poorly known that windows supports IPP, though. On usenet you see lots of folks trying to get windows -> samba -> cups printing to work, and frequently noone seems to know that you could just do windows -> cups directly... :(
My understanding is that the poor delete performance of xfs was a problem with xfs 1.0, and it was fixed in xfs 1.1. The current version, btw, is 1.3.
No doubt you can explain the flaw in their benchmarking process that gave them a factor of 2 speed over sockets.
Factor this or that, but I think it's a clear flaw that he has compared a full-featured production kernel (linux 2.4.20) with his own minimal kernel. It's comparing apples to oranges, IMHO.
Can we expect another alternative out there soon?
Judging from the success of all the other zillion "I'm gonna create something new which will replace X"-projects, I suspect not.
Well since you asked.. :), I find it somewhat strange that these two topics are bundled together like that. The adiabatic approximation is really QM 101-level material, while understanding Berry phase is IMHO quite difficult. And note, it's "Berry phase", not "Berry's phase". The Berry phase approach is probably most famous for it's central part in the modern quantum mechanical theory of polarization. In case you're interested, a review of the topic can be found in e.g.
Resta: "Macroscopic Polarization in Crystalline Dielectrics: The Geometric Phase Approach" Rev. Mod. Phys. 66, 899 (1994)
I'm waiting for the day that Microsoft Windows GUI will be fully raytrace/radiosity/photon map rendered.
Huh, why settle for petty approximations like that. Just to give an incentive to MS, I'm not happy until the light-matter interaction between the light source and the cursor is given a correct quantum mechanical treatment! Of course, with todays computers that means something on the order of a CPU-year for every nanosecond time evolution in a system with 10-100 atoms. But who cares? You know, do something right or don't do it at all!
Um, could you clarify what you mean by "active" vs. "dead" objects?
// setB and getB methods
I meant that with hibernate, if I have a persistent class, say
class Foo {
Bar B;
}
I don't need to explicitly manipulate B through the hibernate API, it works transparently. Eg. if I have loaded an object of the class Foo (say foo), I can access B directly
foo.getB();
Depending on how you configure it, B is loaded at the same time as foo, or it is backed by a proxy and loaded only when it is accessed for the first time.
Of course, being backed by a RDBMS, fancy shit like runtime addition of properties to a class etc. possible with e.g. metaclass hacking in python or MOP in LISP aren't possible. Anyway, if your requirements are on this level, I think Java is the wrong choice for implementation language (aspectj etc. notwithstanding).
Here's a port of prevayler to python: PyPerSyst. The basic concept is rather simple, so I don't think it would be very hard to implement in language X, in case you're interrested.
However, when you have more active objects- which generally arise when you are actually doing good OO design and programming- the interface falls apart some. What if I've got my data objects pointing to other complex objects, rather than just elemental/translatable types like strings, numbers, and dates? What if the state of object A depends directly on what is going on in object B?
In an OORDBMS translation system, the objects are essentially dead while in storage. Object A can point to Object B - another row in a different database which holds data from a different class- but it dead.
In my experience, hibernate solves these kinds of problems quite neatly, with support for collections, dependent objects and whatnot. I'm sure other OR mappers worth anything do the same.
Uh oh, guess how myrinet, quadrics and scali achieve their indeed impressively low latency? By having special user-space MPI libraries that access the hardware directly, without the kernel. And of course, they dont use the IP protocol, but some proprietary protocol designed specifically for cluster use (as simple as possible e.g. no routing)
So, unfortunately the technology used for cluster interconnects is totally non-general purpose. Actually it's more or less useless unless you have a MPI application.
I think the logic goes something like: If you truly love RMS, you aren't using windows anyway.
Its slashdotted right now but IIRC it uses squishdot, a slashdot clone written on top of the zope application server.
I'm a bit sick of reactionary fear of technology. I work for a major university and deal with "outcry" to many of our "potentially dangerous" research projects.
No shit. There was recently a case of an experiment to study vortexes in liquid He or something like that. Anyway, the equations that describe the flow of this fluid is very similar to the equations that describe black holes.. Yeah, you guessed it; some enviro-wackos tried to prohibit the experiment because they were afraid that a black hole would be created which would swallow the earth.
*sigh* indeed.
I hate to tell the reactionaries this, but the people capable of, say, bioengineering plants to extract toxins from the soil, are also the most competent ones for putting in safeguards and policing themselves.
It seems to me that these days the green movement is more about micromanaging peoples lives than about truly caring about the enviroment. I find it very sad.
The world is actually already full of self-replicating nanoscale bots, and at this very moment your very own precious body contains billions of them! Scary, huh? Better outlaw them until they do something dangerous!
They are called "bacteria". They have been around long before us, and they will be here long after the last human has died.
Need to re-read your facts, violent crime has increased as gun ownership has decreased. Conversely in areas of the US where *legal* gun ownership increases, violent crime decreases.
So, where in my post did I say anything that would contradict the above statement? I just pointed out the difference in gun homicide rates.
I didn't say that creating strict gun control laws would magically solve the problem of high gun homicide in the US. In fact, I believe it might not, because an important part of gun control is to make it much harder for criminals to aquire guns. With hundreds of millions of unlicenced guns already circulating in the US, I suspect there might actually be an increase in gun crime because the criminals (who would have no trouble getting guns anyway) would take advantage of the less well armed populace.
I guess that was sort of your point anyway, huh?
You also forgot to mention the appalling reduction of rights in the Britain scenario.
Reduction of what rights? Personally, I'm more afraid of gun nuts shooting at everything that moves during the night because they're afraid of the dark, than of my government turning against me.
And besides, if the government turns against me, what good would it do if I have a handgun and poppas shotgun when they're coming at me with tanks, choppers, GPS-guided bombs and whatnot?
Regardless of the accuracy of your numbers, you actually want to suggest that we trade trade liberty for perceived safety ( and governmental control )?
Well, since I don't live in the USA and have no plans of moving there either, I don't give a flying fuck. If your vision of liberty is that everybody walks around armed to the teeth constantly afraid of each other and the big bad government, so be it. If you're happy with it, good for you.
You sound like a socialist to me...
Because I quoted some figures on gun homicide rates in Britain and the USA? So, everybody that says something that doesn't shine a bright light on the USA is a commie swine? Your vision of liberty doesn't include freedom of thought?
Or, perhaps your extraordinary vision foresaw that I truly am a commie, evil down to the bottom of my rotten heart. BOOH! HISS! Better start digging that foxhole in your backyard before those evil commie tank armies roll you over.
Just look at Britain..
Yup.. the gun homicide rate in Britain is 0.11 per 100,000, while in the US it's 3.72 per 100,000. Go figure.
Umm, yes.. of course. Duh..
Everyone gets a couple of solar panels on the roof for the high-pressure system, and a small turbine for the low-pressure.
Uh, remember economics of scale. A big wind turbine will produce electricity cheaper than many small ones. For solar cell though, your argument is correct. As solar cells get cheaper, the day will certainly come where they will be placed on most roofs. Currently solar power is outrageously expensive though, so it will take a long time.
AC crosses 0 volts 120 (100 in europe) times a second.
Umm, no. In Europe we use 230 V AC (most of Europe at least, perhaps there are exceptions).
I have such a thing in my appartment. And you know what? I never use the low-flush button. Why? Because if I do, the toilet fouls up so fast that you have to clean it twice a week. Ugh. Another factor being that I don't pay a separate water bill, and where I live the water supply is abundant. The water company even has to run water through the mains pipes sometimes to avoid impuritities sticking to the walls of the pipes.
Anyway, as the previous poster said, a more useful system than these low-flush toilets would be to utilize gray water.