For example, if I find that foo.bar.example.com include foo.bar.sourceforge.net GPL'd code, I can blow the whistle on foo.bar.example.com even though I've never had any involvement in foo.bar.sourceforge.net. I've got the same rights to foo.bar.sourceforge.net as the original author, thanks to the GPL.
You can "blow the whistle" all you want, but you can't sue if it's not your code. Thus the original author does have one right that you don't have.
Apple implemented DRM in software, which can be cracked pretty easily. This is because Apple doesn't really care if their DRM is cracked; they just want to do the minimum necessary to keep the RIAA happy. But if the DRM was in hardware, it would potentially be much more difficult for Apple's customers to exercise their fair use rights. So that makes Apple less evil than the MS/Intel/AMD axis.
Windows supports IPv6 already, although not perfectly.
The ThreeDegrees P2P app automatically enables and configures IPv6 when you install it, and all its traffic goes over IPv6. It turned out not to be a killer app, but imagine if something like Kazaa or Skype decided to enable IPv6 on everyone's computer.
The Internet does not support multicast, and will not for the forseeable future, and IPv6 does not change that. The problem is that multicast could create a virtually unlimited amount of state in Internet core routers if it was enabled, but those routers don't have virtually unlimited RAM to hold all that state.
You can "browse" by any attribute, not just filenames. Consider Lifestreams where the "primary" attribute was date/time instead of filename, and all the data appeared as an ordered stream.
Part of the advantage of FP is that it can be automatically parallelized much easier than imperative code, and is easier and less error-prone to write than multithreaded imperative code.
I know that, and I still think most programmers will prefer multithreaded imperative code because it's an incremental change (especially when you throw in OpenMP and Cilk).
Besides, I suspect automatic overlaying is a bigger challenge for Cell than automatic parallelization.
The reason why I mentioned "putting FP back into the curriculum" was, however, that it is my understanding that, if you're right, there's a good chance that programmers would prefer multithreading in imperative languages precisely because it'd be closer to what they'd be used to. So, by getting them used to FP, we'd see a "more fair" evaluation of the practicality of this approach.
Your idea seems to be that idealism would drive the CS curriculum, which then would drive the industry. My observation over the last 10 years has been the opposite. In 1996, CS freshman were learning Scheme and Haskell; today they are using Java, because "it's more practical and aligned with the industry" or some such excuse. But now that the bust has eliminated all the "I just want to get rich" CS students, maybe it will swing the other way.
I think you're reaching for an analogy that isn't there. Intel, AMD, etc. publish specs on their processors, too. This isn't a new thing.
Hardware design is done on computers and chip specifications are more often than not specified in VHDL or Verilog--the "source code" of hardware if you will. Not only is design and simulation within the reach of even hobbyists, the end result is very similar to software in characteristics.
Except when I write software, I can actually run it. If I got ahold of some VHDL for a real chip and modified it... I'd be nowhere since I couldn't actually use the result. (Don't talk to me about FPGAs; they're not competitive with high-volume ASICs.)
Slashdot Mirror
For example, if I find that foo.bar.example.com include foo.bar.sourceforge.net GPL'd code, I can blow the whistle on foo.bar.example.com even though I've never had any involvement in foo.bar.sourceforge.net.
I've got the same rights to foo.bar.sourceforge.net as the original author, thanks to the GPL.
You can "blow the whistle" all you want, but you can't sue if it's not your code. Thus the original author does have one right that you don't have.
Yeah, they are showing the G70 chip and the GeForce 7800.
Apple implemented DRM in software, which can be cracked pretty easily. This is because Apple doesn't really care if their DRM is cracked; they just want to do the minimum necessary to keep the RIAA happy. But if the DRM was in hardware, it would potentially be much more difficult for Apple's customers to exercise their fair use rights. So that makes Apple less evil than the MS/Intel/AMD axis.
This just sounds like a cheaper, built-in version of IPMI which is already included in servers from tier 1 vendors.
Is this LaGrande or something else? Intel promised that they would sell chips both with and without LaGrande; I wonder if they will stick to it.
Intel has a policy of not adding undocumented features to their products, so where's the documentation? Or have they changed their policy?
I'm glad someone got it, although the mods really disappointed me too; I figured it was headed straight for +5, Funny.
Not really. In that world every OS comes with a locked-down firewall by default.
You don't need any ISP support if you use 6to4.
Windows supports IPv6 already, although not perfectly.
The ThreeDegrees P2P app automatically enables and configures IPv6 when you install it, and all its traffic goes over IPv6. It turned out not to be a killer app, but imagine if something like Kazaa or Skype decided to enable IPv6 on everyone's computer.
One problem is that the united states has a lot more IPs per population than most of the rest of the world
Only because ISPs in other countries have not asked for lots of addresses.
we'll be one of the last to run out
Nope; since IP addresses are allocated out of a common worldwide pool, everybody will run out more-or-less at the same time.
Hopefully once all the addresses are allocated they will be privatized so that ISPs can trade address blocks, leading to a more efficient allocation.
Hey, those are difficult graphical effects to do right. I don't think I've ever seen realistic smoke in a console game yet.
Xen will not be included in any Linux kernel, but it's still free and included in the distros so the result is more or less the same.
IPv6 addresses are assigned using the same process as IPv4.
IPv6 guarrantees nothing, since "mandatory to implement" and "mandatory to enable" are quite different things (and neither one is enforceable anyway).
The Internet does not support multicast, and will not for the forseeable future, and IPv6 does not change that. The problem is that multicast could create a virtually unlimited amount of state in Internet core routers if it was enabled, but those routers don't have virtually unlimited RAM to hold all that state.
IPSec works the same for IPv4 and IPv6. The same.
(Don't even mention "mandatory to implement", since there is no enforcement.)
...a custom tool designed to fit their hands perfectly and accompany them in the way that they (the developers) work...
That's what BitKeeper is, too. The genesis of BK was a meeting between Linus and Larry many years ago.
You can "browse" by any attribute, not just filenames. Consider Lifestreams where the "primary" attribute was date/time instead of filename, and all the data appeared as an ordered stream.
Xenon was the codename for the whole Xbox 360; the processor chip has a different codename (at least within IBM).
Now it's nano-clicking. The sound is so small you can't even hear it.
Part of the advantage of FP is that it can be automatically parallelized much easier than imperative code, and is easier and less error-prone to write than multithreaded imperative code.
I know that, and I still think most programmers will prefer multithreaded imperative code because it's an incremental change (especially when you throw in OpenMP and Cilk).
Besides, I suspect automatic overlaying is a bigger challenge for Cell than automatic parallelization.
The hypervisor and Linux both run on the PPE. The SPUs only run application code.
I hope you're joking, since Java programmers still ought to understand those basic concepts.
I don't get it. My Web site and (incoming) mail server is not hosted by my ISP, but I happily send my outgoing mail through my ISP's SMTP server.
The reason why I mentioned "putting FP back into the curriculum" was, however, that it is my understanding that, if you're right, there's a good chance that programmers would prefer multithreading in imperative languages precisely because it'd be closer to what they'd be used to. So, by getting them used to FP, we'd see a "more fair" evaluation of the practicality of this approach.
Your idea seems to be that idealism would drive the CS curriculum, which then would drive the industry. My observation over the last 10 years has been the opposite. In 1996, CS freshman were learning Scheme and Haskell; today they are using Java, because "it's more practical and aligned with the industry" or some such excuse. But now that the bust has eliminated all the "I just want to get rich" CS students, maybe it will swing the other way.
I think you're reaching for an analogy that isn't there. Intel, AMD, etc. publish specs on their processors, too. This isn't a new thing.
Hardware design is done on computers and chip specifications are more often than not specified in VHDL or Verilog--the "source code" of hardware if you will. Not only is design and simulation within the reach of even hobbyists, the end result is very similar to software in characteristics.
Except when I write software, I can actually run it. If I got ahold of some VHDL for a real chip and modified it... I'd be nowhere since I couldn't actually use the result. (Don't talk to me about FPGAs; they're not competitive with high-volume ASICs.)