I don't think size is an issue really. Faster cycling doesn't come from adding transistors, it comes from making things happen faster. If anything, putting things closer together helps.
If you think such a weapon is not under development you are fooling yourself.
Uh... Right. Several issues there.
We can build something as extraordinarily powerful as a nuclear weapon because there is a lot of energy to be released from the fission of uranium or plutonium or whatever. This energy is stored in the bonds inside the unstable nuclei, and we just let it out. It was originally put in there when some exploding star made the uranium nuclei in question, long before the solar system was formed. We do not have to provide that energy.
Thing is, there are not similar reserves of naturally-occurring antimatter to be mined, because... well, it's kind of obvious. The problem is this: current (and any sensible-sounding future) methods of antimatter production involve actually putting in at least the amount of energy you want to get out. That mass won't come out of nowhere you know. So while it's all fine to say that an antimatter weapon would be scary because a really really small one could knock the planet off course, I have to ask you where you you think we're gonna get that much energy from. Maybe from a nuclear power plant? The amount of uranium used is going to be the amount you'd need to make a normal nuke big enough to do the same job, isn't it (that is to say, more than could conceivably be acquired)?
Also, what makes you think that the threat of total annihilation would bring peace? The threat of total annihilation is here already. Russia and the USA maintain far more weapons than they need to completely destroy the other, partly as protection against missiles being hit while still on the ground, etc. If all the world's weapons were to be detonated, it would likely destroy human life on the planet. If such a thing as a world-destroying antimatter bomb ever existed, people would do what the Soviet Union and the USA did with their nuclear arsenals: basically agree not to use them, and go on fighting with conventional weapons (yes, I know they didn't officially fight each other at any point, but USSR armed the Vietcong, US armed the Mujahideen, etc.).
The 7700 has a somewhat strange way of controlling fan speed. Irritatingly, it cannot be controlled through software like some fans, but instead has a little box with a variable resistor adjustable by a little knob, with an adhesive patch to mount it outside the case. Maybe it's because it's harder to silence a pulse width modulated fan than one which is regulated by a constant voltage.
I have a CNPS7700-AlCu. It's a cheaper one than the one they review - and also smaller. The piece of metal with the vanes sits straight on the CPU with no heat pipes or anything. What they don't mention is that even my smaller cooler is technically out-of-spec - they're heavier than a LGA775 cooler should be, but motherboards don't actually snap that easily.
Anyway, the cooler comes with a device for adjusting it's speed, and it is practically silent on the lowest setting while still providing pretty good cooling. It helps that my processor isn't a very hot one (Intel Core 2 Duo 6300), but even on the silent setting I cannot make it go over 49 C. In fact, the vanes have enough surface area that if it's a cold day, the cooler works fine disconnected, i.e. without the fan turning.
As they say, fitting it can be a pain, but that is presumably the price you pay for fitting some 700g of copper on the motherboard.
By the way, it's worth taking measurements or checking their list of supported motherboards - it's physical dimensions are beyond the LGA775 spec as well. It extends out over the components immediately surrounding the CPU, and on my motherboard it neatly blows air through the northbridge and GPU heatsinks.
It would be more fun to make this do some arbitrary software thing. You could do this by wiring it to the power button or something (on ACPI systems under Linux, at least, it isn't hard to make that do whatever you want - just have a program watching/proc/acpi/event).
It might be fun to bind it to "skip track" and hit it hard whenever that song you meant to delete ages ago comes up.
You CAN render a whole page using one TCP/IP connection already!
It's called HTTP pipelining. Basically, the browser is allowed to send another request after it finishes recieving content, using the same connection. By the time the actual HTML has finished being transfered, it should be ready to ask for the images and CSS. Doing things that way reduces the impact of latency. At least Opera, Firefox (if you enable it in about:config) and Apache support this.
(In fact AFAIK all servers should support it if they claim to be HTTP 1.1. ISS doesn't support it, of course.)
Where is this new search? Usually Google doesn't announce things till they are ready (well, or at least beta;-). One of the things I've always like about Google is that it seldom builds up some product/vapourware before it's release. Is that policy going to end now?
I can play Q3A-based games, at least, with Beryl running. There is not a significant reduction in FPS. The "Unredirect fullscreen windows" option helps, but my screensaver (Euphoria) looks pretty good even without that.
Back when I used XGL with old Nvidia drivers, I did indeed have problems with FPS in OpenGL apps, because there was basically no way to render directly. I used to have a script to launch games in a second X session (and I still use it for games which are difficult to minimise - it's nice to still be able to respond to IMs). Now I do not use XGL or AIGLX, just Xorg and recent Nvidia drivers. This system has no such limitation.
However, I currently have Beryl set up to make all windows of type "unknown" semi-transparent, which covers qt menus and various strange OSDs (e.g. Amarok's), and generally looks very nice but sadly also covers fullscreen OpenGL windows. I hope there is a workaround for this soon, as many KDE users have things set up like that.
So ATM I do what many people who have problems with Beryl do: I just start another WM before running a game, and start Beryl again afterwards. Once you've made the shortcuts to do this, it's no trouble at all.
All in all, Beryl should be considered still in devel and you might have small problems but they are generally easy to fix or quickly work around.
I appreciate Google using an open standard for their IM system, as I use Linux and have been able to use it despite the absence of a Linux client from Google, as it works with any IM client supporting Jabber.
However, who says that their video extension to the protocol will be an open standard?
I have seen and heard of several people using old/AOL CDs to scare birds away from various vegetable gardens and allotments. You hang CDs from a tree/fence/whatever (or along a string between to stakes) by a thread through the centre, and the constantly moving rainbow patterns as the CDs spin in the wind confuse and scare birds. They reportedly work much better than scarecrows (most types of bird soon work out that scarecrows can't really move and just get used to them).
Slashdot Mirror
I don't think size is an issue really. Faster cycling doesn't come from adding transistors, it comes from making things happen faster. If anything, putting things closer together helps.
I think that might the the definition of a litre, not of a kg.
Please, please, please mod parent underrated. This is the first post I've ever seen which deserves +5, Offtopic.
We can build something as extraordinarily powerful as a nuclear weapon because there is a lot of energy to be released from the fission of uranium or plutonium or whatever. This energy is stored in the bonds inside the unstable nuclei, and we just let it out. It was originally put in there when some exploding star made the uranium nuclei in question, long before the solar system was formed. We do not have to provide that energy.
Thing is, there are not similar reserves of naturally-occurring antimatter to be mined, because... well, it's kind of obvious. The problem is this: current (and any sensible-sounding future) methods of antimatter production involve actually putting in at least the amount of energy you want to get out. That mass won't come out of nowhere you know. So while it's all fine to say that an antimatter weapon would be scary because a really really small one could knock the planet off course, I have to ask you where you you think we're gonna get that much energy from. Maybe from a nuclear power plant? The amount of uranium used is going to be the amount you'd need to make a normal nuke big enough to do the same job, isn't it (that is to say, more than could conceivably be acquired)?
Also, what makes you think that the threat of total annihilation would bring peace? The threat of total annihilation is here already. Russia and the USA maintain far more weapons than they need to completely destroy the other, partly as protection against missiles being hit while still on the ground, etc. If all the world's weapons were to be detonated, it would likely destroy human life on the planet. If such a thing as a world-destroying antimatter bomb ever existed, people would do what the Soviet Union and the USA did with their nuclear arsenals: basically agree not to use them, and go on fighting with conventional weapons (yes, I know they didn't officially fight each other at any point, but USSR armed the Vietcong, US armed the Mujahideen, etc.).
The 7700 has a somewhat strange way of controlling fan speed. Irritatingly, it cannot be controlled through software like some fans, but instead has a little box with a variable resistor adjustable by a little knob, with an adhesive patch to mount it outside the case. Maybe it's because it's harder to silence a pulse width modulated fan than one which is regulated by a constant voltage.
I have a CNPS7700-AlCu. It's a cheaper one than the one they review - and also smaller. The piece of metal with the vanes sits straight on the CPU with no heat pipes or anything. What they don't mention is that even my smaller cooler is technically out-of-spec - they're heavier than a LGA775 cooler should be, but motherboards don't actually snap that easily.
Anyway, the cooler comes with a device for adjusting it's speed, and it is practically silent on the lowest setting while still providing pretty good cooling. It helps that my processor isn't a very hot one (Intel Core 2 Duo 6300), but even on the silent setting I cannot make it go over 49 C. In fact, the vanes have enough surface area that if it's a cold day, the cooler works fine disconnected, i.e. without the fan turning.
As they say, fitting it can be a pain, but that is presumably the price you pay for fitting some 700g of copper on the motherboard.
By the way, it's worth taking measurements or checking their list of supported motherboards - it's physical dimensions are beyond the LGA775 spec as well. It extends out over the components immediately surrounding the CPU, and on my motherboard it neatly blows air through the northbridge and GPU heatsinks.
It would be more fun to make this do some arbitrary software thing. You could do this by wiring it to the power button or something (on ACPI systems under Linux, at least, it isn't hard to make that do whatever you want - just have a program watching /proc/acpi/event).
It might be fun to bind it to "skip track" and hit it hard whenever that song you meant to delete ages ago comes up.
What bug is this?
You CAN render a whole page using one TCP/IP connection already!
It's called HTTP pipelining. Basically, the browser is allowed to send another request after it finishes recieving content, using the same connection. By the time the actual HTML has finished being transfered, it should be ready to ask for the images and CSS. Doing things that way reduces the impact of latency. At least Opera, Firefox (if you enable it in about:config) and Apache support this.
(In fact AFAIK all servers should support it if they claim to be HTTP 1.1. ISS doesn't support it, of course.)
Ooh! Wasn't doing that last time I looked! I wonder if it is being rolled out by region, or by random cookies like the last change was.
Where is this new search? Usually Google doesn't announce things till they are ready (well, or at least beta ;-). One of the things I've always like about Google is that it seldom builds up some product/vapourware before it's release. Is that policy going to end now?
Last time I checked, MSN's file transfer sucked with every client, including the official one.
It sounds like a pretty scary catch-all if it includes writing essays. what else is considered "disorderly conduct" under US law?
Also, doesn't the US have a constitution which makes freedom of expression an absolute right?
We're not objecting to that.
We're laughing at how low the demand must be.
I can play Q3A-based games, at least, with Beryl running. There is not a significant reduction in FPS. The "Unredirect fullscreen windows" option helps, but my screensaver (Euphoria) looks pretty good even without that.
Back when I used XGL with old Nvidia drivers, I did indeed have problems with FPS in OpenGL apps, because there was basically no way to render directly. I used to have a script to launch games in a second X session (and I still use it for games which are difficult to minimise - it's nice to still be able to respond to IMs). Now I do not use XGL or AIGLX, just Xorg and recent Nvidia drivers. This system has no such limitation.
However, I currently have Beryl set up to make all windows of type "unknown" semi-transparent, which covers qt menus and various strange OSDs (e.g. Amarok's), and generally looks very nice but sadly also covers fullscreen OpenGL windows. I hope there is a workaround for this soon, as many KDE users have things set up like that.
So ATM I do what many people who have problems with Beryl do: I just start another WM before running a game, and start Beryl again afterwards. Once you've made the shortcuts to do this, it's no trouble at all.
All in all, Beryl should be considered still in devel and you might have small problems but they are generally easy to fix or quickly work around.
Umm... Turn off that plugin? Beryl is pretty modular.
I appreciate Google using an open standard for their IM system, as I use Linux and have been able to use it despite the absence of a Linux client from Google, as it works with any IM client supporting Jabber.
However, who says that their video extension to the protocol will be an open standard?
MacOS support came a long time after Windows support. Linux support still isn't done and isn't generally talked about.
I think they *are* choking on it: surely no one actually has the current bid of "$1e+12". :-)
Do they not have some kind of defence against blatant trolling, given that that's really what this auction is about?
Hopefully this will let Skype start quietly dieing the way it should have when it started providing video support for Windows clients only...
Can you boot any burned disks? Did you use a CD-RW (not a CD-R) in a drive which doesn't support those?
According to Wikipedia, Penryn is intended as a laptop processor.
Does it seem odd to anyone else for Intel to launch a new instruction set on a laptop CPU? Are portables that dominant these days?
I have seen and heard of several people using old/AOL CDs to scare birds away from various vegetable gardens and allotments. You hang CDs from a tree/fence/whatever (or along a string between to stakes) by a thread through the centre, and the constantly moving rainbow patterns as the CDs spin in the wind confuse and scare birds. They reportedly work much better than scarecrows (most types of bird soon work out that scarecrows can't really move and just get used to them).
In other news, I call for a '.stupid' domain.