But don't use misleading attributions. The first sentence is Hitler, from Mein Kampf. It was speaking on the view that the duty of the people is to produce healthy children and not burden society with the support of children. Not to protect the children, but to have useful children. A disturbing sentiment when considering how extreme Hitler took things like this, but orthogonal to this discussion.
Rabbi Daniel Lapin is the person who actually wrote that quote, putting the totalitarian twist on it to link it to an excuse to curtail liberty. It's insightful, but not directly linked to Hitler's strategy for totalitarianism. He wasn't nearly so subtle as that.
He did explicitly call out unicast, which youtube is doing. He's not advocating the likes of youtube, or even saying this isn't an improvement. He's saying that networking explicitly designed multicast as an architecture for 'getting it right'. On a theoretical level, I have to see his point, on a more pragmatic level, good luck fixing all the stuff in the middle instead of just the endpoints..
First off 'broadcast' packets have poor meaning outside of a single layer2 network. In networking, broadcast means to send to everyone, whether they like it or not.
I presume you meant multicast, which is a bit more sane, but at the same time, I haven't been convinced that an at-scale situation would work with networks generally as configured today. I wager a good number of arbitrary routers out there would fubar multicast. This, of course, doesn't get rid of the duplicates, it moves some of the problem to the networking backbone a bit and still you end up with retries for the clients that missed pieces. For streaming, it's particularly interesting if a client misses a packet. Generally, when not dealing with real-time considerations, multicast clients save what they can, then at the end ask for pieces they missed. With streaming, that is going on all the time as a client can't wait til the ill-defined 'end' to be orderly.
Now, with all the networking infrastructure being pretty unicast tolerant, and not guaranteed much else, the p2p stuff makes some sense. Are there some inefficiencies on the host side, probably, but unicast is dead simple for networks to get correct.
The fact of the matter is neither nVidia nor ATI is used to dealing with the pace of linux changes. This is made more complicated by the reality of trying to support so many distributions at different points in that spectrum while trying to patch your own bugs. Doing this while at the same time preserving the relatively straightforward task of presenting a simple, single, versioned package is staggering.
As a result, frequently they have problems and non-compliance issues. Right this second, for example, the closed drivers don't play perfectly well with xrandr, and you have to use proprietary tools to change multihead configuration.
As much as I love linux as a user, some of the design decisions make it really hard for one vendor to support multiple distributions. If they open source it, then each distribution of note will make the effort to bring it in and customize it.
The way you speak, it's like you install the latest vintage of the popular vintages of distributions and ATI hardware should magically have drivers loaded.
It also makes it sounds like the 3D goodness of the OSS driver is mature and even working nearly immediately after spec release.
It also suggests that the ATI binary blob has not been plagued with problems.
The simple fact is, the 3D support isn't still fully baked even in the source control trees, especially not in any distribution. The effort to run those drivers is not insurmountable, but it does require checking out code, compiling it, etc etc. It's not in something that people would declare a release. Note even demo, proof-of-concept glxgears on one system was actually a couple of months after the specs released.
And while the binary driver has markedly improved, you still have high incidents of 'unsupported hardware' watermarks, some corruption, suspend problems, and more depending on hardware. Speaking from experience as a Radeon driven laptop that has been chasing the code. Yes, the 3D performance finally got respectable, and AIGLX finally got implemented, but it's still quite rough around the edges.
That said, I hope by Intrepid in ubuntu world and Fedora 10 in that world, they include OSS ATI goodness, as those drivers are really coming into their own on their development trees.
Quit simply, if it is a small, close community and you want to use SSL, have them add your and only your certificate, with plenty of ways to make sure it's fine.
The problem with CACert, is that you are extending your statement to the greater internet, not just yourself. Adding the certificate for a website is one thing, adding a CA certificate of an organization whose validation process is not as thorough (for good reason) is another.
CACert might be good *if* the browser but a bar indicating the lower barrier of entry to get one. cacert would be more than adequate for most forums, mailing list archives, and non-financial account information on sites you don't have much to lose on. For financial transactions, I would want only to participate with a site that went through a more rigorous validation process than CACerts. Such processes require manual human attention to do right, which isn't free.
I think fab is a significant problem, but ultimately, at this point, AMD's offering catching up is insufficient to be interesting. Going above and beyond (like when they brought about x86-64) is really what they need right now.
IBM buys AMD, uses circumstances to: -Advance the fab capabilities of AMD generally (hopefully invest to actually keep up with Intel instead of lagging by a year or so) -Release a Cell processor variant, replacing the PPC core with an x86 core.
It seems far fetched, but at the same time, the #1 supercomputer is already an AMD/Cell hybrid (two Cell processor packages for every AMD package). However, I wouldn't anticipate that core being any more performance than the PPC core, just a different instruction set. It *could* really cause some grief for intel if it caught on though. The ability to run Windows and games like normal (maybe with a penalty), but SPU enabled software could really make for some amazing media manipulation and incredible games.
As it stands, it's pretty dire. The question is, can AMD turn around and match the 45nm process with a decent design before the Nehalem generation? I wonder that explicitly because the last bragging point they have is their interprocessor architecture and memory controller, which Nehalem matches. If Intel releases that and the rest of AMD's tech remains as disadvantaged as it is, watch for some of the 4-socket and above space that AMD still has some sway in move to Intel.
I can see your point, and it probably won't be until the October/November timeframe at best before distributions will make current-gen AMD/ATI graphics have 3D out of the box in an OSS way.
I personally used nVidia recently, though this laptop is AMD with their binary driver, which has been improving at least.
How would have AMD impeded use of good 3D cards? Even if you thought nVidia SLI was the only 'good' answer, there are nForce chipsets for AMD with SLI too... I personally don't buy into the price-power-performance ratio of SLI or CrossFire, btw.
It is in part semantics, but at the same time it also represents the core not ostensibly having a bugfix-only branch. Distributions fill in the gap there to an extent. But it does reflect a departure from a lot of common practice of having a branch to follow for those content with featuresets, but needing the security and bug fixes too. As of the no-2.7 branch, this was already the case, this truly is just semantics. But the 2.7 decision was about more than semantics.
The firmware update fiasco was horrible, but it's not bricked. Very few things that people call 'bricked' really is. Swap in a borrowed hard drive, update to 2.41, swap in old hard drive, problem solved. It's not easy, and by no means excusable, but it isn't 'bricked'. It's also better than a hardware failure that 360s were historically plagued with.
The crossplatform games can be blamed on developers probably more than the raw capability. The PS3 GPU is probably better situated, and the Cell can churn out more raw horsepower. The problem being that to acheive that, the PS3 is a far less flexible platform. In 360, the memory can be tuned to the games needs to have x amount of main memory and y for gpu. With PS3, it's always a 50/50 split (though the memory is faster). With the PPC core, the lower amount of computation power in 360 is aggregated, thus easier to leverage. In PS3, you've got to know how and be able to divide your workload into 6 pieces to take advantage. Even FFXIII stated they were only scaling to 4.
For my part, I just want competition to continue. Nintendo showed back in the day what happens when a company has too much control over an industry, and thankfully they mis-stepped horribly in putting N64 against Playstation.
I run an 8800GT off a decent 380W power supply. The power on the 12V line is abnormally high for a 380W rating, but still. The 8800GT does require an extra connector. My Antec Solo keeps it respectable at medium fanspeed on the single 12 CM fan. I pummelled it repeatedly over the months and could not get it to hang or do anything erratic, so I'm confident that this power supply is adequate for my setup.
A problem pervading power supply 'requirements', is that no vendor can require that simple rating. The actual requirement is more along the lines of 'this device will need X AMPs on this particular 12V DC circuit, gather all the requirements and sum them for each rail/voltaige'. Instead of doing this, they say 'need 500W power supply', as that gives them some headroom for crappy power supplies and a flexibility of choices of other components. This is one example of why a system OEMer *could* potentially do better than a home builder, as they can extract the hard data from the vendors more readily, and size a power supply correctly for the components instead of having to pull out the overkill parts.
The metadata is signed as well as all the packages. This mitigates the situation, i.e. a malicious mirror cannot host arbitrary combination of packages to suit their needs, they must mirror some permutation that I had at one point approved all together.
It's never gotten full. 25TB would be no problem. Perhaps our technologies could combine. Linux even has a driver for it. You can access it using/dev/null.
The read-only device has worked great for me too. I've only ever gotten 0s out of it./dev/zero and/dev/null combined could be awesome.
In this case wouldn't your new friend encourage you to think of a tinfoil hat as ludicrous? I mean, after all, what is in your friend's best interests...
What have we to lose from allowing felons the right to vote? Surely they won't override the proportion of the country that is against the activity if it is unjust? Do you honestly think murderer's will find a 'pro-murder' candidate to vote for, and if so, do you honestly think people at large would vote for such a candidate? Even among convicted murderers, I doubt you'd find many wanting to legalize that act, considering the likely shortened lifespan at the hands of those who would exact revenge within the extent of the law.
If a law puts a majority share of people into jail/gives them a felony, there is almost certainly something wrong with the law. A good and just law should have no worries of being challenged by the masses, even those who have violated current sets of laws.
Firearms may also be a bit dicey of a question given the existence of non-violent felonies that in and of themselves aren't related to firearms ownership/use. However I think the original intent of the second amendment is moot in this day and age anyway, as an ultimate check and balance against the government. If a true just rebellion would have to occur, military vehicles/weapons would have to be subverted without overt force before such a rebellion could have a chance anyway. The difference between the firearms a citizen is allowed to have and what the military controls is too great.
To save money, the RIAA then asks for a dismissal without prejudice.
Given the scale of RIAA legal costs incurred during that prolonged phase, I would say they aren't trying to save money at all. The defendant's legal fees are trivial next to the cost of trying to drag it out to avoid paying them.
It's quite obvious they want everyone to feel that regardless of the findings of a court, the defendant will have lost, and it's best to do as they say in the first place, regardless of guilt or innocence. RIAA doesn't care much about appearing to be in the right or anything anymore, they just want to induce fear, safe in the knowledge people will still fund the music industry as they way RIAA leads it.
Given the obvious systematic tactics of making a business of milking the Justice system, the courts really need to thoroughly strike down RIAA's efforts.
Slashdot Mirror
But don't use misleading attributions. The first sentence is Hitler, from Mein Kampf. It was speaking on the view that the duty of the people is to produce healthy children and not burden society with the support of children. Not to protect the children, but to have useful children. A disturbing sentiment when considering how extreme Hitler took things like this, but orthogonal to this discussion.
Rabbi Daniel Lapin is the person who actually wrote that quote, putting the totalitarian twist on it to link it to an excuse to curtail liberty. It's insightful, but not directly linked to Hitler's strategy for totalitarianism. He wasn't nearly so subtle as that.
He did explicitly call out unicast, which youtube is doing. He's not advocating the likes of youtube, or even saying this isn't an improvement. He's saying that networking explicitly designed multicast as an architecture for 'getting it right'. On a theoretical level, I have to see his point, on a more pragmatic level, good luck fixing all the stuff in the middle instead of just the endpoints..
First off 'broadcast' packets have poor meaning outside of a single layer2 network. In networking, broadcast means to send to everyone, whether they like it or not.
I presume you meant multicast, which is a bit more sane, but at the same time, I haven't been convinced that an at-scale situation would work with networks generally as configured today. I wager a good number of arbitrary routers out there would fubar multicast. This, of course, doesn't get rid of the duplicates, it moves some of the problem to the networking backbone a bit and still you end up with retries for the clients that missed pieces. For streaming, it's particularly interesting if a client misses a packet. Generally, when not dealing with real-time considerations, multicast clients save what they can, then at the end ask for pieces they missed. With streaming, that is going on all the time as a client can't wait til the ill-defined 'end' to be orderly.
Now, with all the networking infrastructure being pretty unicast tolerant, and not guaranteed much else, the p2p stuff makes some sense. Are there some inefficiencies on the host side, probably, but unicast is dead simple for networks to get correct.
The fact of the matter is neither nVidia nor ATI is used to dealing with the pace of linux changes. This is made more complicated by the reality of trying to support so many distributions at different points in that spectrum while trying to patch your own bugs. Doing this while at the same time preserving the relatively straightforward task of presenting a simple, single, versioned package is staggering.
As a result, frequently they have problems and non-compliance issues. Right this second, for example, the closed drivers don't play perfectly well with xrandr, and you have to use proprietary tools to change multihead configuration.
As much as I love linux as a user, some of the design decisions make it really hard for one vendor to support multiple distributions. If they open source it, then each distribution of note will make the effort to bring it in and customize it.
The way you speak, it's like you install the latest vintage of the popular vintages of distributions and ATI hardware should magically have drivers loaded.
It also makes it sounds like the 3D goodness of the OSS driver is mature and even working nearly immediately after spec release.
It also suggests that the ATI binary blob has not been plagued with problems.
The simple fact is, the 3D support isn't still fully baked even in the source control trees, especially not in any distribution. The effort to run those drivers is not insurmountable, but it does require checking out code, compiling it, etc etc. It's not in something that people would declare a release. Note even demo, proof-of-concept glxgears on one system was actually a couple of months after the specs released.
And while the binary driver has markedly improved, you still have high incidents of 'unsupported hardware' watermarks, some corruption, suspend problems, and more depending on hardware. Speaking from experience as a Radeon driven laptop that has been chasing the code. Yes, the 3D performance finally got respectable, and AIGLX finally got implemented, but it's still quite rough around the edges.
That said, I hope by Intrepid in ubuntu world and Fedora 10 in that world, they include OSS ATI goodness, as those drivers are really coming into their own on their development trees.
Quit simply, if it is a small, close community and you want to use SSL, have them add your and only your certificate, with plenty of ways to make sure it's fine.
The problem with CACert, is that you are extending your statement to the greater internet, not just yourself. Adding the certificate for a website is one thing, adding a CA certificate of an organization whose validation process is not as thorough (for good reason) is another.
CACert might be good *if* the browser but a bar indicating the lower barrier of entry to get one. cacert would be more than adequate for most forums, mailing list archives, and non-financial account information on sites you don't have much to lose on. For financial transactions, I would want only to participate with a site that went through a more rigorous validation process than CACerts. Such processes require manual human attention to do right, which isn't free.
The sad part is, that thought didn't even occur to me at the time.
Skipping ahead a bit in the logical progression...
&&&&&&&<
Ok, I officially went and made it unfunny now.
I think fab is a significant problem, but ultimately, at this point, AMD's offering catching up is insufficient to be interesting. Going above and beyond (like when they brought about x86-64) is really what they need right now.
Though I doubt it would ever happen.
IBM buys AMD, uses circumstances to:
-Advance the fab capabilities of AMD generally (hopefully invest to actually keep up with Intel instead of lagging by a year or so)
-Release a Cell processor variant, replacing the PPC core with an x86 core.
It seems far fetched, but at the same time, the #1 supercomputer is already an AMD/Cell hybrid (two Cell processor packages for every AMD package). However, I wouldn't anticipate that core being any more performance than the PPC core, just a different instruction set. It *could* really cause some grief for intel if it caught on though. The ability to run Windows and games like normal (maybe with a penalty), but SPU enabled software could really make for some amazing media manipulation and incredible games.
As it stands, it's pretty dire. The question is, can AMD turn around and match the 45nm process with a decent design before the Nehalem generation? I wonder that explicitly because the last bragging point they have is their interprocessor architecture and memory controller, which Nehalem matches. If Intel releases that and the rest of AMD's tech remains as disadvantaged as it is, watch for some of the 4-socket and above space that AMD still has some sway in move to Intel.
I can see your point, and it probably won't be until the October/November timeframe at best before distributions will make current-gen AMD/ATI graphics have 3D out of the box in an OSS way.
I personally used nVidia recently, though this laptop is AMD with their binary driver, which has been improving at least.
How would have AMD impeded use of good 3D cards? Even if you thought nVidia SLI was the only 'good' answer, there are nForce chipsets for AMD with SLI too... I personally don't buy into the price-power-performance ratio of SLI or CrossFire, btw.
by typing <
did you do just < or did you do <?
My Mythbox uses suspend to RAM. If you pick the right components, Linux will have no issues.
It is in part semantics, but at the same time it also represents the core not ostensibly having a bugfix-only branch. Distributions fill in the gap there to an extent. But it does reflect a departure from a lot of common practice of having a branch to follow for those content with featuresets, but needing the security and bug fixes too. As of the no-2.7 branch, this was already the case, this truly is just semantics. But the 2.7 decision was about more than semantics.
The firmware update fiasco was horrible, but it's not bricked. Very few things that people call 'bricked' really is. Swap in a borrowed hard drive, update to 2.41, swap in old hard drive, problem solved. It's not easy, and by no means excusable, but it isn't 'bricked'. It's also better than a hardware failure that 360s were historically plagued with.
The crossplatform games can be blamed on developers probably more than the raw capability. The PS3 GPU is probably better situated, and the Cell can churn out more raw horsepower. The problem being that to acheive that, the PS3 is a far less flexible platform. In 360, the memory can be tuned to the games needs to have x amount of main memory and y for gpu. With PS3, it's always a 50/50 split (though the memory is faster). With the PPC core, the lower amount of computation power in 360 is aggregated, thus easier to leverage. In PS3, you've got to know how and be able to divide your workload into 6 pieces to take advantage. Even FFXIII stated they were only scaling to 4.
For my part, I just want competition to continue. Nintendo showed back in the day what happens when a company has too much control over an industry, and thankfully they mis-stepped horribly in putting N64 against Playstation.
I run an 8800GT off a decent 380W power supply. The power on the 12V line is abnormally high for a 380W rating, but still. The 8800GT does require an extra connector. My Antec Solo keeps it respectable at medium fanspeed on the single 12 CM fan. I pummelled it repeatedly over the months and could not get it to hang or do anything erratic, so I'm confident that this power supply is adequate for my setup.
A problem pervading power supply 'requirements', is that no vendor can require that simple rating. The actual requirement is more along the lines of 'this device will need X AMPs on this particular 12V DC circuit, gather all the requirements and sum them for each rail/voltaige'. Instead of doing this, they say 'need 500W power supply', as that gives them some headroom for crappy power supplies and a flexibility of choices of other components. This is one example of why a system OEMer *could* potentially do better than a home builder, as they can extract the hard data from the vendors more readily, and size a power supply correctly for the components instead of having to pull out the overkill parts.
The metadata is signed as well as all the packages. This mitigates the situation, i.e. a malicious mirror cannot host arbitrary combination of packages to suit their needs, they must mirror some permutation that I had at one point approved all together.
It seems to be a bad copy of Second Life.
That's a severe accusation. I tried Second Life. I thought of it as all the design 'quality' and intelligence of myspace, now with 3D goodness...
It's never gotten full. 25TB would be no problem. Perhaps our technologies could combine. Linux even has a driver for it. You can access it using /dev/null.
The read-only device has worked great for me too. I've only ever gotten 0s out of it. /dev/zero and /dev/null combined could be awesome.
In this case wouldn't your new friend encourage you to think of a tinfoil hat as ludicrous? I mean, after all, what is in your friend's best interests...
What have we to lose from allowing felons the right to vote? Surely they won't override the proportion of the country that is against the activity if it is unjust? Do you honestly think murderer's will find a 'pro-murder' candidate to vote for, and if so, do you honestly think people at large would vote for such a candidate? Even among convicted murderers, I doubt you'd find many wanting to legalize that act, considering the likely shortened lifespan at the hands of those who would exact revenge within the extent of the law.
If a law puts a majority share of people into jail/gives them a felony, there is almost certainly something wrong with the law. A good and just law should have no worries of being challenged by the masses, even those who have violated current sets of laws.
Firearms may also be a bit dicey of a question given the existence of non-violent felonies that in and of themselves aren't related to firearms ownership/use. However I think the original intent of the second amendment is moot in this day and age anyway, as an ultimate check and balance against the government. If a true just rebellion would have to occur, military vehicles/weapons would have to be subverted without overt force before such a rebellion could have a chance anyway. The difference between the firearms a citizen is allowed to have and what the military controls is too great.
To save money, the RIAA then asks for a dismissal without prejudice.
Given the scale of RIAA legal costs incurred during that prolonged phase, I would say they aren't trying to save money at all. The defendant's legal fees are trivial next to the cost of trying to drag it out to avoid paying them.
It's quite obvious they want everyone to feel that regardless of the findings of a court, the defendant will have lost, and it's best to do as they say in the first place, regardless of guilt or innocence. RIAA doesn't care much about appearing to be in the right or anything anymore, they just want to induce fear, safe in the knowledge people will still fund the music industry as they way RIAA leads it.
Given the obvious systematic tactics of making a business of milking the Justice system, the courts really need to thoroughly strike down RIAA's efforts.