Since when do reputable software vendors just drop support for their product out of the blue without any notice or suitable upgrade/replacement?
There are plenty of suitable replacements. ReiserFS was becoming irrelevant anyway, it had scalabilty issues and Hans seemed to have lost interest in fixing it. If a free/OSS project that really is crucial stalls, then invariably it gets picked up. For an example look at how X.org forked off XFree86, and that wasn't even "dead" it just had a ridiculously protracted development process.
In a way RISC is still very much alive, it's just confined to the internals of the chips these days, ie microcode. When internal processing speeds started to outstrip external bus speeds, the simplicity of RISC became a disadvantage as it increases the amount of external fetch operations. Complex instruction sets allow more of a "work package" to be delivered to the processor to work on within it's fast internal cache.
Here in the UK we use old-fashioned paper ballots, hand counted. No tabulating machines, no hanging chads, no technology at all. In a General Election, the polls close at 10PM and the earliest constituencies usually declare their results around 1AM. By 8AM the next morning there are only a few left to declare and the result is known. This is in a country of some 60 million people - there is no reason why it couldn't scale up to the US population.
Why complicate things and introduce more potential for fraud?
Graphics on Server should perform just as well as on Player. If you find they don't, you may not be making correct use of the "VMWare Tools", which install on the guest OS and provide optimised graphics drivers for the virtual graphics adapter. In any case, a virtualized interface is never going to be ideal. The best solution is to ignore the VM console and use whatever remote admin facilities the Guest OS has. I run XP Pro under VMWare Server and connect via RDP. This gives pretty much native performance when running locally. You can do the same thing with player, but you're stuck with a useless VM console window you don't need.
Completely missed the point of my post. I'm not interested in discussing the particular controversies over Iraq, I'm making a point about the validity of the "Greater good" argument and whether it constitutes moral relativism.
For example, claiming that slaughtering civilians is unconscionable when Saddam Hussein does it, but is in the service of a greater good when we do it, is a fine example of moral relativism.
Poor example.
Saddam Hussein slaughtered innocent civilians deliberately. Civilians casualties in wars happen as a side effect of action taken in the pursuit of genuine military objectives. Deliberate targetting of civilians is a war crime. Untill we invent the magic bomb that kills the bad guys, leaving the innocent unharmed, the "greater good" argument seems valid to me. Whether what we did in Iraq was a "greater good" is of course open to considerable debate, but try applying that logic to a less controversial conflict. Innocent civilians were undoubtably killed by Allied bombs in WWII. Were we wrong to fight? Was that a morally relativist thing to do? It all comes down to intent and justification, which is why I made sure to include those points in my original statement.
Anti-abortion ideas along with other values that are mostly based on religion can't really be debated because they are based on dogma and not on reality or reason. Although anti-abortionists sometimes try to tie their ideas with reality it's a futile excersise and their arguments always end up being about faith and endoctrinated religious beliefs.
You shouldn't discount an entire position based on the poor advocacy skills of some of those who support it. Anti-abortion ideas are based on a moral framework. For the majority of pro-lifers this framework is based on their faith, which unfortunately leads many to resort to relgious arguments which boil down to "because God says so", sure to lose any argument with a non-believer. The irony is that the basic moral underpinning of the pro-life case is not at all exclusive to religious people, it is shared by almost everyone, and that is:
It's wrong to deliberately kill an innocent person without justification.
Most people accept this. Then the debate becomes about what constitutes a deliberate act, a person and justification.
Of course, not everyone accepts this. If you don't acknowledge that we should have a universal moral framework of some sort, you're a moral relativist and can't say anything is wrong. Was it wrong what those guys did on September 11th? Well, it depends on your point of view...
Incidentally, if you want to get really hot about abortion, take The Pill off the market. The old high-dosage Pill used to work by stopping ovulation. That dosage had too many side-effects. The new low-dosage Pill, basically the only kind in use now, works by preventing implantation. Essentially the low-dosage Pill is a very early (clearly pre-brain) abortion.
Your conclusion is correct, but you're a little off on the facts. The original high dose pill works as you describe, the second description you gave is how the morning after (or "emergency contraceptive" pill works (and why pro-lifers oppose it). The modern low dose contraceptive pill is something of a halfway house between the two. It's primary action is still to stop ovulation in the same way as the original pill, but reducing the dose made the failure rate too high. To remediate this a second hormone was introduced to provide a backup anti-implantation action. Therefore a woman taking the modern pill over a long period of time may have some of these early abortions without even knowing about it.
Frankly, with abortion, I think medical science is going to resolve the issue for us in about 30 years (maybe less), certainly before the issue can be solved by the partisan debate. Every woman will have the right to evict a baby from her womb at any time, and that baby will have the right to the best medical care available during and after the procedure, which will essentially mean that any woman who wants not to carry a baby to term will have the baby transferred to an artificial womb, with the costs born by the state. She'll give up all parental rights by doing so, and the resulting baby will have no legal connection with her whatsoever.
This "solution" presupposes that the purpose of women having abortions is merely to avoid the remaining months of their pregnancy. In every other sense it would have an identical outcome to carrying to term and giving up for adoption, which has always been available. When choosing abortion over adoption women generally rank the inconvenience of the pregnancy itself pretty low on their list of considerations. The most commonly given reasons are:
1. The anguish of knowing that your child is out there, never being able to have contact or even know how it is doing.
2. The notion that at it's current stage of development it is just a "blob of tissue" or "clump of cells", not yet a "person" and so not such a terrible thing to dispose of.
To go to the effort of constructing an artificial womb to save the fetus would be to acknowledge the falsehood of reason 2, and leave reason 1 unresolved.
In addition to these, there are lots of complex "rights" arguments made by pro-choice activists to say that development is irrelevant, even if we acknowlege the personhood of the fetus the mother still has the right to evict it from her body. However, these are rarely cited by the women who actually have abortions. It's usually very important to them to regard the fetus as being inconsequential. Unsurprising really, as being told you're "entitled" to kill an innocent is unlikely to make you feel any better about it. The pro-life side make counter-arguments that the baby has the right to be there. (Which IMO make a heck of a lot more sense, but then I am a pro-life activist.)
Note that these arguments generally apply to the majority of abortions for "social" reasons, not those for fetal abnormality (there's a whole load of extra arguments in there).
Granted, disk capacity is growing faster than tape capacity can keep up, but RAID cannot be an offsite bacup solution.
And here lies the problem. Well, that and the fact that the drives and tapes cost waaay too much for a home user to consider. Cost per GB for the tape cartridges may be cheaper than disks, but you have to be using a hell of a lot of tapes to offset the cost of the drive. Which probably means a library system... which costs even more. For a datacenter, yes.
For a home user, more disks are the only economical option. A simple mirror image of their data is enough for most people. Incremental point-in-time backups may serve to protect users from themselves (like realizing you shouldn't have deleted that vital document three weeks ago), but they're not really necessary if your main concern is hardware failure. When I screw up, I usually realize immediately so can retrieve from the last backup.
I agree with you assement of RAID, and that offsite is best, but for a home user you also have to consider what is practical. An on-site-but-as-far-away-as-possible solution that runs automatically over the network every day is better than offsite backups that don't get done.
I require 1TB of fully backed up network storage and I've been researching my own solution. My thoughts:
1.) None of the "home" NAS appliances came close to meeting my requirements and the pro setups are too expensive. Building my own boxes is the way to go.
2.) I looked at SATA raid 5 cards, but then decided to forget the redundant aspect of RAID. Raid 1/3/5 are for uptime, they enable a properly configured server to hot-swap failed disks without rebooting, not important in a home setup. I realised how useless it is for backup if, say, the PSU were to overvolt and fry all the disks simultaneously. I'm still considering RAID 0 for performance though.
3.) Following on from above, anything in the same case is vulnerable to a catastrophic failure. To properly back things up, I need two boxes sited as far away from each other as possible.
3.) I want decent performance and I've got a lot of data to move across the network. Gig ethernet is obviously the way to go. Gig ethernet NICs struggle over PCI, so I want either a PCI-E NIC, or preferably an intergrated motherboard solution that bypasses the PCI bus.
4.) To maximise future expansion potential, I want motherboards with lots of SATA ports and big disks so I don't take up too many of them.
The ethernet throughput and SATA port requirements led me straight to NForce 3/4. I'm going for Nforce 4 as the price difference is not too great and PCI-E with socket 939 gives better future upgrade potential. (I've already decided to use the boxes for more than just storage, who knows what I'll want to do with them in a year's time.) If you're certain you'll only use them as pure network storage, then NForce 3/socket 754 and the cheapest Sempron should easily suffice.
I'm going for two Seagate 500GB disks per box. The cost per gig is marginally higher than mid-capacity disks (optimum seems about 300GB at the moment), but the benefits of future expandabilty outweighed that for me. If I want to expand further in future, I can easily take it to 2TB by moving all four into one box and purchasing 2TB more in whatever disk capacity is available/economical at that time. The 5 year guarantee won me over to Seagate.
To ensure backup is constantly maintained with the minimum amount of effort and network congestion, I will be running rsync between the two boxes. I'm also looking at the possibily of a dedicated backup network between them by using dual gig-ethernet motherboards. As I will have full data redundancy between the The two boxes, I figure I have nothing to lose striping the data between the two disks in each box to maximise throuput performance.
The setup I'm going for should, theoretically at least, come close to maxing out gig ethernet throughput and be as fast (in data tranfer terms, don't know about latency) as a typical locally connected hard drive. However, none of this is implemented yet, so if anyone can see a glaring error, please tell me before I waste money!
Slashdot Mirror
There are plenty of suitable replacements. ReiserFS was becoming irrelevant anyway, it had scalabilty issues and Hans seemed to have lost interest in fixing it. If a free/OSS project that really is crucial stalls, then invariably it gets picked up. For an example look at how X.org forked off XFree86, and that wasn't even "dead" it just had a ridiculously protracted development process.
In a way RISC is still very much alive, it's just confined to the internals of the chips these days, ie microcode. When internal processing speeds started to outstrip external bus speeds, the simplicity of RISC became a disadvantage as it increases the amount of external fetch operations. Complex instruction sets allow more of a "work package" to be delivered to the processor to work on within it's fast internal cache.
You have a votes that need to be counted. What difference does the system of goverment make?
Here in the UK we use old-fashioned paper ballots, hand counted. No tabulating machines, no hanging chads, no technology at all. In a General Election, the polls close at 10PM and the earliest constituencies usually declare their results around 1AM. By 8AM the next morning there are only a few left to declare and the result is known. This is in a country of some 60 million people - there is no reason why it couldn't scale up to the US population. Why complicate things and introduce more potential for fraud?
Graphics on Server should perform just as well as on Player. If you find they don't, you may not be making correct use of the "VMWare Tools", which install on the guest OS and provide optimised graphics drivers for the virtual graphics adapter. In any case, a virtualized interface is never going to be ideal. The best solution is to ignore the VM console and use whatever remote admin facilities the Guest OS has. I run XP Pro under VMWare Server and connect via RDP. This gives pretty much native performance when running locally. You can do the same thing with player, but you're stuck with a useless VM console window you don't need.
Completely missed the point of my post. I'm not interested in discussing the particular controversies over Iraq, I'm making a point about the validity of the "Greater good" argument and whether it constitutes moral relativism.
For example, claiming that slaughtering civilians is unconscionable when Saddam Hussein does it, but is in the service of a greater good when we do it, is a fine example of moral relativism.
Poor example.
Saddam Hussein slaughtered innocent civilians deliberately. Civilians casualties in wars happen as a side effect of action taken in the pursuit of genuine military objectives. Deliberate targetting of civilians is a war crime. Untill we invent the magic bomb that kills the bad guys, leaving the innocent unharmed, the "greater good" argument seems valid to me. Whether what we did in Iraq was a "greater good" is of course open to considerable debate, but try applying that logic to a less controversial conflict. Innocent civilians were undoubtably killed by Allied bombs in WWII. Were we wrong to fight? Was that a morally relativist thing to do? It all comes down to intent and justification, which is why I made sure to include those points in my original statement.
Anti-abortion ideas along with other values that are mostly based on religion can't really be debated because they are based on dogma and not on reality or reason. Although anti-abortionists sometimes try to tie their ideas with reality it's a futile excersise and their arguments always end up being about faith and endoctrinated religious beliefs.
You shouldn't discount an entire position based on the poor advocacy skills of some of those who support it. Anti-abortion ideas are based on a moral framework. For the majority of pro-lifers this framework is based on their faith, which unfortunately leads many to resort to relgious arguments which boil down to "because God says so", sure to lose any argument with a non-believer. The irony is that the basic moral underpinning of the pro-life case is not at all exclusive to religious people, it is shared by almost everyone, and that is:
It's wrong to deliberately kill an innocent person without justification.
Most people accept this. Then the debate becomes about what constitutes a deliberate act, a person and justification.
Of course, not everyone accepts this. If you don't acknowledge that we should have a universal moral framework of some sort, you're a moral relativist and can't say anything is wrong. Was it wrong what those guys did on September 11th? Well, it depends on your point of view...
Incidentally, if you want to get really hot about abortion, take The Pill off the market. The old high-dosage Pill used to work by stopping ovulation. That dosage had too many side-effects. The new low-dosage Pill, basically the only kind in use now, works by preventing implantation. Essentially the low-dosage Pill is a very early (clearly pre-brain) abortion.
Your conclusion is correct, but you're a little off on the facts. The original high dose pill works as you describe, the second description you gave is how the morning after (or "emergency contraceptive" pill works (and why pro-lifers oppose it). The modern low dose contraceptive pill is something of a halfway house between the two. It's primary action is still to stop ovulation in the same way as the original pill, but reducing the dose made the failure rate too high. To remediate this a second hormone was introduced to provide a backup anti-implantation action. Therefore a woman taking the modern pill over a long period of time may have some of these early abortions without even knowing about it.
Frankly, with abortion, I think medical science is going to resolve the issue for us in about 30 years (maybe less), certainly before the issue can be solved by the partisan debate. Every woman will have the right to evict a baby from her womb at any time, and that baby will have the right to the best medical care available during and after the procedure, which will essentially mean that any woman who wants not to carry a baby to term will have the baby transferred to an artificial womb, with the costs born by the state. She'll give up all parental rights by doing so, and the resulting baby will have no legal connection with her whatsoever.
This "solution" presupposes that the purpose of women having abortions is merely to avoid the remaining months of their pregnancy. In every other sense it would have an identical outcome to carrying to term and giving up for adoption, which has always been available. When choosing abortion over adoption women generally rank the inconvenience of the pregnancy itself pretty low on their list of considerations. The most commonly given reasons are:
1. The anguish of knowing that your child is out there, never being able to have contact or even know how it is doing.
2. The notion that at it's current stage of development it is just a "blob of tissue" or "clump of cells", not yet a "person" and so not such a terrible thing to dispose of.
To go to the effort of constructing an artificial womb to save the fetus would be to acknowledge the falsehood of reason 2, and leave reason 1 unresolved.
In addition to these, there are lots of complex "rights" arguments made by pro-choice activists to say that development is irrelevant, even if we acknowlege the personhood of the fetus the mother still has the right to evict it from her body. However, these are rarely cited by the women who actually have abortions. It's usually very important to them to regard the fetus as being inconsequential. Unsurprising really, as being told you're "entitled" to kill an innocent is unlikely to make you feel any better about it. The pro-life side make counter-arguments that the baby has the right to be there. (Which IMO make a heck of a lot more sense, but then I am a pro-life activist.)
Note that these arguments generally apply to the majority of abortions for "social" reasons, not those for fetal abnormality (there's a whole load of extra arguments in there).
Mount the drives flat, not on edge.
Why? Modern drives are (allegedly) supposed to work fine in either orientation.
Granted, disk capacity is growing faster than tape capacity can keep up, but RAID cannot be an offsite bacup solution.
And here lies the problem. Well, that and the fact that the drives and tapes cost waaay too much for a home user to consider. Cost per GB for the tape cartridges may be cheaper than disks, but you have to be using a hell of a lot of tapes to offset the cost of the drive. Which probably means a library system... which costs even more. For a datacenter, yes.
For a home user, more disks are the only economical option. A simple mirror image of their data is enough for most people. Incremental point-in-time backups may serve to protect users from themselves (like realizing you shouldn't have deleted that vital document three weeks ago), but they're not really necessary if your main concern is hardware failure. When I screw up, I usually realize immediately so can retrieve from the last backup.
I agree with you assement of RAID, and that offsite is best, but for a home user you also have to consider what is practical. An on-site-but-as-far-away-as-possible solution that runs automatically over the network every day is better than offsite backups that don't get done.
I require 1TB of fully backed up network storage and I've been researching my own solution. My thoughts:
1.) None of the "home" NAS appliances came close to meeting my requirements and the pro setups are too expensive. Building my own boxes is the way to go.
2.) I looked at SATA raid 5 cards, but then decided to forget the redundant aspect of RAID. Raid 1/3/5 are for uptime, they enable a properly configured server to hot-swap failed disks without rebooting, not important in a home setup. I realised how useless it is for backup if, say, the PSU were to overvolt and fry all the disks simultaneously. I'm still considering RAID 0 for performance though.
3.) Following on from above, anything in the same case is vulnerable to a catastrophic failure. To properly back things up, I need two boxes sited as far away from each other as possible.
3.) I want decent performance and I've got a lot of data to move across the network. Gig ethernet is obviously the way to go. Gig ethernet NICs struggle over PCI, so I want either a PCI-E NIC, or preferably an intergrated motherboard solution that bypasses the PCI bus.
4.) To maximise future expansion potential, I want motherboards with lots of SATA ports and big disks so I don't take up too many of them.
The ethernet throughput and SATA port requirements led me straight to NForce 3/4. I'm going for Nforce 4 as the price difference is not too great and PCI-E with socket 939 gives better future upgrade potential. (I've already decided to use the boxes for more than just storage, who knows what I'll want to do with them in a year's time.) If you're certain you'll only use them as pure network storage, then NForce 3/socket 754 and the cheapest Sempron should easily suffice.
I'm going for two Seagate 500GB disks per box. The cost per gig is marginally higher than mid-capacity disks (optimum seems about 300GB at the moment), but the benefits of future expandabilty outweighed that for me. If I want to expand further in future, I can easily take it to 2TB by moving all four into one box and purchasing 2TB more in whatever disk capacity is available/economical at that time. The 5 year guarantee won me over to Seagate.
To ensure backup is constantly maintained with the minimum amount of effort and network congestion, I will be running rsync between the two boxes. I'm also looking at the possibily of a dedicated backup network between them by using dual gig-ethernet motherboards. As I will have full data redundancy between the The two boxes, I figure I have nothing to lose striping the data between the two disks in each box to maximise throuput performance.
The setup I'm going for should, theoretically at least, come close to maxing out gig ethernet throughput and be as fast (in data tranfer terms, don't know about latency) as a typical locally connected hard drive. However, none of this is implemented yet, so if anyone can see a glaring error, please tell me before I waste money!