Would it really see massive savings if it used VoIP over their own network? ATM packets are still 48 bit, having to have to throw another network protocol in there would just increase the overhead. And if they don't gurantee a channel, there's always a chance that someone's packet is going to be dropped if they load up lines with more than however many 48k channels would fit.
Technically, you should be able to put it in the.cpp file, as per Stroustrup's book. gcc does properly implement templates (I should say, does not fully implement), but if you use something like Sun or Microsoft's compiler, it works correctly. Note that this is not an endorsement of Microsoft's compiler.
You can use class introspection in Java to find out its proper class, it's part of the polymorphism by dynamic binding of runtime procedures design of the VM.
Or, you could use Objective-C, ala NeXTSTEP, where the generic pointer id always knows that the object type, and therefor never downcasts to the root object (Java was based in part on Objective-C). The nice thing about Objective-C is that unhandled exceptions aren't volatile (well, unless you end up trying derefrence a null pointer). In either case, yes, C++ templates are probably a great degree faster, but then again you loose all the dynamic features of a proper runtime system (so then you'd have to declare your methods virtual, which would end up causing your C++ app to run slower, but with the added pain of templates when you're using the STL).
The worst part about Java is its masochistic typechecking system. It's got all of these nice runtime features, but the compile-time typechecker really stands in your way as far as accessing this backend power.
This would be good for soda, such as Code Red, which contains Brominated Vegetable Oil, a chemical on the FDA's watch list as a potentially poisonous compound.
Well, no, IPv6 addresses are never supposed to be statically allocated, instead relying completely on hostname resolution. Not to mention the fact that you probably couldn't remember an IPv6 easily, not to mention several. If this conforms to the same in-arpa DNS system, then this would imply that the enum numbers would provide a reverse lookup mapping. I don't know anything about enum though, so this is all conjecture.
Counter-balanced by the multiple heirarchy caches, though. Most x86 architecture machines don't have an L3 cache. The mere fact that there's a 2MB L3 cache should make the processor a lot faster. In fact, Apple probably doesn't even actually need DDR memory onboard, it's probably just for show. I'm willing to bet that the bottleneck is *not* at the memory bus.
CISC only does more per cycle if you assume that it takes one cycle to complete every instruction. This is not the case. Intel chips take CISC code, break it down into their own microcode, which is then processed. The clock speed that Intel reports is the clock speed of the chip that processes this microcode. The smaller the instructions can be divided up, the faster they can run the chip at. Clock speed is not necessarily a basis for comparing two different chips, even when of the same architecture, especially when Intel is involved.
Why is there no minimum number of colors for a 3D graph? I can't think of an example that would require an infinite number of colors, therefor the statement about there not being a minimum number of colors appears to be false.
I don't understand how this can just work, you may be able to swap for a yellow or green LED, but a blue LED has a much higher activation voltage, which would mean that either the red LED was being overdriven, or there's a chip in charge of changing the voltage somehow because it was designed to handle other color LEDs as well. The blue LED should be really dim if the voltage is for a red LED (GaAS).
Upon further research into the Danger Hiptop, it appears that it has a java vm. Therefor you should be able to run any java telnet application, such as java-ssh http://www.cl.cam.ac.uk/~fapp2/software/java-ssh/ or whatever. HTH.
You forgot to mention the fact that the 7455, while having a one cycle longer fp unit, has its three fp cores separated so data can be fetched to all three at the same time. And that the altivec units are also separated into four distinct types allowing several different types of instructions to be loaded and processed in parallel.
The Apple HIG strictly state that a Mac OS X application should never sit behind the dock. If you look at all of Apple's apps, they will automatically resize their windows if you change the vertical size of the dock. Applications that maximize to behind the dock are in violation of Apple's HIG, another reason that a lot of ports of windows applications are considered crappy.
I don't know what you guys are talking about, the DivX codec from divx.com works extremely well on my PBG4 500MHz, and their translator is very fast, producing files that are about 1-2MB larger than PC files, but play with PC codecs as well for DivX. Really the only problem with playing normal DivX files is that quicktime can't handle variable bit rate MP3s because the data is spread out across AVI frames which is out of spec, thus apple isn't/doesn't support it.
Well, it's not coming out of apple, it's from a third party. It's basically the same thing that has been out for several years on the PC (on the fly MPEG-1 converter, such as the one made by Dazzle), with the addition of scheduled recording and a TV guide. I don't think MPEG-1 requires that much bandwidth, Dazzle made a device that transferred data over the parallel port!
So if we had an anti-spam law in place, would the MPAA be able to be sued under the fact that it was e-mail from someone we had never done business with in the past and had not requested the e-mail?
I was there too, hopefully when this new release is available the Alloy Orchestra will be allowed to license it. I greatly prefer the Alloy Orchestra's soundtrack to that of the original. I can't wait for this movie release to make it to DVD though, the rest of the DVD releases of Metropolis have horrible video quality!
If you're planning on mastering a DVD video, wouldn't your battery die before the process is complete (encoding + burning)? If so, then what's the point? How long does it take to burn a DVD anyway?
Slashdot Mirror
Would it really see massive savings if it used VoIP over their own network? ATM packets are still 48 bit, having to have to throw another network protocol in there would just increase the overhead. And if they don't gurantee a channel, there's always a chance that someone's packet is going to be dropped if they load up lines with more than however many 48k channels would fit.
Technically, you should be able to put it in the .cpp file, as per Stroustrup's book. gcc does properly implement templates (I should say, does not fully implement), but if you use something like Sun or Microsoft's compiler, it works correctly. Note that this is not an endorsement of Microsoft's compiler.
You can use class introspection in Java to find out its proper class, it's part of the polymorphism by dynamic binding of runtime procedures design of the VM.
Or, you could use Objective-C, ala NeXTSTEP, where the generic pointer id always knows that the object type, and therefor never downcasts to the root object (Java was based in part on Objective-C). The nice thing about Objective-C is that unhandled exceptions aren't volatile (well, unless you end up trying derefrence a null pointer). In either case, yes, C++ templates are probably a great degree faster, but then again you loose all the dynamic features of a proper runtime system (so then you'd have to declare your methods virtual, which would end up causing your C++ app to run slower, but with the added pain of templates when you're using the STL).
The worst part about Java is its masochistic typechecking system. It's got all of these nice runtime features, but the compile-time typechecker really stands in your way as far as accessing this backend power.
This would be good for soda, such as Code Red, which contains Brominated Vegetable Oil, a chemical on the FDA's watch list as a potentially poisonous compound.
Well, no, IPv6 addresses are never supposed to be statically allocated, instead relying completely on hostname resolution. Not to mention the fact that you probably couldn't remember an IPv6 easily, not to mention several. If this conforms to the same in-arpa DNS system, then this would imply that the enum numbers would provide a reverse lookup mapping. I don't know anything about enum though, so this is all conjecture.
Yeah, and then spammers from other states add your e-mail address to their list...
Counter-balanced by the multiple heirarchy caches, though. Most x86 architecture machines don't have an L3 cache. The mere fact that there's a 2MB L3 cache should make the processor a lot faster. In fact, Apple probably doesn't even actually need DDR memory onboard, it's probably just for show. I'm willing to bet that the bottleneck is *not* at the memory bus.
CISC only does more per cycle if you assume that it takes one cycle to complete every instruction. This is not the case. Intel chips take CISC code, break it down into their own microcode, which is then processed. The clock speed that Intel reports is the clock speed of the chip that processes this microcode. The smaller the instructions can be divided up, the faster they can run the chip at. Clock speed is not necessarily a basis for comparing two different chips, even when of the same architecture, especially when Intel is involved.
Why is there no minimum number of colors for a 3D graph? I can't think of an example that would require an infinite number of colors, therefor the statement about there not being a minimum number of colors appears to be false.
I don't understand how this can just work, you may be able to swap for a yellow or green LED, but a blue LED has a much higher activation voltage, which would mean that either the red LED was being overdriven, or there's a chip in charge of changing the voltage somehow because it was designed to handle other color LEDs as well. The blue LED should be really dim if the voltage is for a red LED (GaAS).
He was even better as Ernest Pratt and Nicodemus Legend in the UPN show, Legend.
Alexey Pajitnov.
Upon further research into the Danger Hiptop, it appears that it has a java vm. Therefor you should be able to run any java telnet application, such as java-ssh http://www.cl.cam.ac.uk/~fapp2/software/java-ssh/ or whatever. HTH.
Someone could fake an aim packet though, make it appear as if rm -rf / were coming from your screenname, and that would suck.
You might want to consider OTP (one time passwords) for login, so you avoid typing your password plaintext.
They're not actually one time use, they can be used for up to one month from the time they are issued as many times as you want.
You forgot to mention the fact that the 7455, while having a one cycle longer fp unit, has its three fp cores separated so data can be fetched to all three at the same time. And that the altivec units are also separated into four distinct types allowing several different types of instructions to be loaded and processed in parallel.
The Apple HIG strictly state that a Mac OS X application should never sit behind the dock. If you look at all of Apple's apps, they will automatically resize their windows if you change the vertical size of the dock. Applications that maximize to behind the dock are in violation of Apple's HIG, another reason that a lot of ports of windows applications are considered crappy.
It doesn't work anyway, even if you do have a three button mouse on OS X.
I don't know what you guys are talking about, the DivX codec from divx.com works extremely well on my PBG4 500MHz, and their translator is very fast, producing files that are about 1-2MB larger than PC files, but play with PC codecs as well for DivX. Really the only problem with playing normal DivX files is that quicktime can't handle variable bit rate MP3s because the data is spread out across AVI frames which is out of spec, thus apple isn't/doesn't support it.
How is this any different than what Web Objects does? Or is it not different?
Well, it's not coming out of apple, it's from a third party. It's basically the same thing that has been out for several years on the PC (on the fly MPEG-1 converter, such as the one made by Dazzle), with the addition of scheduled recording and a TV guide. I don't think MPEG-1 requires that much bandwidth, Dazzle made a device that transferred data over the parallel port!
Well, I hate to break it to you, but this is not a rumor. See: http://www.elgato.com/eyeTV/index.html for more details.
So if we had an anti-spam law in place, would the MPAA be able to be sued under the fact that it was e-mail from someone we had never done business with in the past and had not requested the e-mail?
I was there too, hopefully when this new release is available the Alloy Orchestra will be allowed to license it. I greatly prefer the Alloy Orchestra's soundtrack to that of the original. I can't wait for this movie release to make it to DVD though, the rest of the DVD releases of Metropolis have horrible video quality!
If you're planning on mastering a DVD video, wouldn't your battery die before the process is complete (encoding + burning)? If so, then what's the point? How long does it take to burn a DVD anyway?