Rasterization is JUST a cheap trick to make it look something like that, nothing else. Right, but it's really fast. It's faster than ray tracing is good.
If you're waiting for humans to get rid of fast approximations when they're good enough, I hope you're patient.
I'd much rather have volume or block level snapshots, like with LVM and other similar systems. Those systems provide RO and RW snapshots, dynamic partitioning, drive spanning, etc., and can be easily layered with other block-level components to provide compression, encryption, remote storage, etc. as well. All that without tying you to a single file system (though that may be a moot point on OS X, as it will only boot from HFS/HFS+ AFAIK). ZFS shits all over LVM:
-Say I want to take hourly snapshots, and retain them for a month. When the parent data for a ZFS snapshot changes, ZFS merely has to leave the old data alone. OTOH, LVM must copy the block to every snapshot before it can change it in the parent. My hourly snapshots will quickly cause my disk to thrash to a halt with LVM and using much more space, while ZFS incurs a negligible penalty.
-LVMs allow dynamic partitioning, but they can't share capacity on the fly. If I delete a file on an LVM-hosted filesystem, that space becomes available to the filesystem but not all the others. Unless I shrink the filesystem, generally requiring that I take it offline for a while.
-Another layer could potentially handle checksums on LVMs, but in practice Linux can't do this properly by itself.
-ZFS can use other layers, there's just a substantial benefit to letting it run the show.
The only reason this won't turn out to be a huge disadvantage for Linux is that BTRFS will provide most of the same features. Layering can be a very helpful design tool, but there are times it becomes a hinderence. It's important to be flexible when there's benefits to integrating stuff into a single layer.
ZFS uses super-paranoidal checksumming which can detect drive problems in advance. It won't detect them in advance. But, used appropriately, you won't care when they happen.
I'm not sure that's relevant to single-drive Macs though. It needs a mirror with a clean copy of the data to correct from.
I think most people would be okay with ZFS if the license were amenable to inclusion in the Linux kernel, but right now the fact of the matter is that it's not. FUSE isn't acceptable for core functionality, so basically that means Linux doesn't have the option of using ZFS as a successor to any of the major 4 filesystems. If Sun were to change their minds about the license that would quickly change, but Sun picked the license with the specific intent of remaining incompatible with Linux so I would regard that as pretty unlikely.
But, even with Reiser gone I'm not sure the impact on reiserfs3 will be immediate. His company abandoned maintenance on it long ago, I think it's handled by someone in Suse now. It's in a precarious situation anyway, and Reiser would not address that.
Honestly though, I'm not sure the maintenance situation for XFS or JFS is so much better that they solve the problem. If you want an assurance of very good support long term, ext3 is basically your only choice.
BTRFS is the one bright spot on the horizon that I can see. It offers many of the same features as ZFS. It's likely going to be years before it's production ready, but if/when that happens it'll be enough better than the existing Linux filesystems to be the clear successor, and I think it'll be close enough to ZFS that most Linux users will stop pining for ZFS support.
I'm guessing the reason MS hasn't done this already is their heavy reliance on undocumented API calls and all the little hidden "tweaks" MS puts into their OS's to make sure MS apps run OK: "IF App=Word97,THEN [some custom module no one knows about except a programmer that isn't there anymore]. And I'm pretty sure there are a shipload of those. So by the time you have created the emulation layer, you have basically recreated the older OS. If it were done by means of virtualization, they could probably retain most of of the old OS to do the work for them.
First, yes NIN's making a lot more because of all the attention this has received, but the amount of attention won't be 'zero' if this becomes mainstream. Major internet sites are not the lower bound for word of mouth.
Second, for every big name like NIN that stands to make millions, there's got to be orders of magnitude more people that can make a decent living. I would be very skeptical that this approach goes from millions in a few days to almost nothing for everyone else.
Third, for bands too obscure to make much money doing this, would they really be making a whole lot via traditional CD sales? They'll still have to start out playing seedy bars.
Of course, if you're hand-compiling everything, you won't have time to apply a patch every time an exploit is discovered in something you do need. You'll be less secure unless you have implausibly large amounts of time to track and maintain everything.
It's not blindly using vendor binaries, there's nothing blind about it. It's a very explicit cost/benefit analysis. Those vendor binaries save time upfront with the install, they save time with patches and upgrades, they vastly increase the number of servers an admin can maintain, and they make support with whatever vendors you're dealing with much easier.
Most here want neutral ISPs. The only way we're going to get that is if the ISP makes money either way, and their "unlimited" plans aren't the way to do that. They were built around assumptions of usage that are no longer valid with bittorrent, iTunes, YouTube, etc being as big as they are now.
One response is to throttle services that consume too many resources. The other is to bill according to usage and upgrade as necessary. That way, Time Warner has a business case to roll out their upgrades and facilitate higher usage, regardless of what you're doing with it.
OTOH, if they see revenue from every gigabyte you download, then they have an incentive to facilitate your usage, rather than transparently throttling it whenever you use a service with an inconveniently large network footprint. As more people download more content, their infrastructure takes a beating with the unlimited plans they now have trouble providing. Per-gigabyte pricing means they have an incentive to upgrade rather than throttle, because there's money in it for them.
Consolidation of ownership notwithstanding, ISPs built their infrastructure and "unlimited" plans around assumptions of usage that are no longer valid. This would be a problem even in a healthy, competitive market. The push to tiered QoS with bandwidth hogs like iTunes and YouTube being throttled is, at least in part, a response to this. ISPs don't want to expose users to the costs of their usage, so they're trying to blackmail it out of everyone else. From a network neutrality perspective, per-gigabyte pricing is great, because it means Time Warner wins whether I'm downloading from iTunes, YouTube, BitTorrrent, etc. It doesn't matter to them because they can simply bill me according to my usage (or, I'll cut back, or, I'll switch to another ISP and take my heavy usage with me).
With pebble bed reactors (and other newer designs), they are "passively safe". This basically means that when the thermal energy in the core gets high enough, it is no longer a neutron-friendly environment and fission can't be sustained. The reactor is designed such that heat is naturally lost to the environment fast enough to keep the equilibrium temperature below that which would melt anything in the core. A meltdown in that case would be impossible, because heat by itself acts like emergency control rods, no operators or computers have to get anything right to halt the reaction.
Erm... it was safe. The reactor suffered a significant failure of its cooling system without escalating beyond that. No design that relies on zero accidents is safe, that requires a design where even accidents aren't catastrophic.
If the sodium leak had involved sodium exposed directly to the reactor (it didn't), environmental contamination would primarily be from Sodium-24, which has a half-life of 15 hours. That's not irreversible contamination, at worst that's a temporary evacuation order. It's not like non-nuclear industrial accidents can't get that bad (eg, the Bhopal disaster). This is something we already have to live with as an industrialized civilization, nuclear power simply has the virtue of releasing far less CO2.
The only part of this that causes me concern is the cover up.
If they've advertised the service as being "unlimited", they have an incentive transparently throttle people that test the limits of their infrastructure.
If the pricing is based on however many dollars per gigabyte, then they have an incentive to facilitate my usage, not throttle it.
The network in your area may not be that heavily loaded, but that's a big problem in more and more areas, and ISPs are worried about the costs of upgrading their networks. Passing the costs of heavy usage on to the consumer generates the revenue needed to justify upgrades to the network. Or, the user cuts back. Or, they switch to another ISP and take their above average usage with them. I'm fine with all the above.
Figure out what it's going to cost to support my usage, bill me accordingly, and upgrade the infrastructure as necessary. I'll either cut back or bear the cost of my higher usage.
Unlike many posters here I actually travel very little. That means I'm not prepared to spend much on a laptop. But, I've found the laptops in my price range tend to be 8+ lb monstrosities. I've got my desktop for heavy lifting, but an Eee is the best choice for very portable and low cost at the same time.
The specs aren't much, but they're enough to get you into Windows XP or a full desktop Linux. You're running the same applications with the peripherals of your choice etc. That's a huge asset.
This tradeoff is not a good one for everyone, but for those of us in a position to benefit, it's the best laptop in years.
"The question is how complex is complex enough? The only criteria we know is enough, is interaction with an intelligent observer, a.k.a. observation. Because observation is the only way we can determine the outcome."
Okay. Since the universe is more complex than any of the intelligent observers therein, this shouldn't be a problem by virtue of the fact that the light from the phenomena we're observing would have hit a great many objects without our intervention. We're not dooming the universe with our hubris by catching the light a few nanoseconds before it would have hit the ground anyway.
> It's not clear to me what problem is being solved by having virtual OSX.
The change to their license only applies to OS X server, so it it applies to their server product. I run my domains on a $20/mo VPS service. This is fine for dns, web, e-mail, moderate database, etc. A server like this is more than enough for a small business.
Without support for virtualization, I'd need to buy and colocate an XServe, which would be around an order of magnitude more expensive. This pretty much rules OS X out without even getting into a discussion of the pros and cons, it just can't participate at a price that's interesting.
"As far as I recall, BSD is basically "do what you want as long as the attribution remains""
No. The first clause in a 3 clause BSD license is "Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer, without modification.".
Precisely. Stating bandwidth limits and overage charges upfront is not "unlimited", but sneaking in a bunch of throttling and limitations lets them continue to claim it's unlimited.
Slashdot Mirror
If you're waiting for humans to get rid of fast approximations when they're good enough, I hope you're patient.
-Say I want to take hourly snapshots, and retain them for a month. When the parent data for a ZFS snapshot changes, ZFS merely has to leave the old data alone. OTOH, LVM must copy the block to every snapshot before it can change it in the parent. My hourly snapshots will quickly cause my disk to thrash to a halt with LVM and using much more space, while ZFS incurs a negligible penalty.
-LVMs allow dynamic partitioning, but they can't share capacity on the fly. If I delete a file on an LVM-hosted filesystem, that space becomes available to the filesystem but not all the others. Unless I shrink the filesystem, generally requiring that I take it offline for a while.
-Another layer could potentially handle checksums on LVMs, but in practice Linux can't do this properly by itself.
-ZFS can use other layers, there's just a substantial benefit to letting it run the show.
The only reason this won't turn out to be a huge disadvantage for Linux is that BTRFS will provide most of the same features. Layering can be a very helpful design tool, but there are times it becomes a hinderence. It's important to be flexible when there's benefits to integrating stuff into a single layer.
I'm not sure that's relevant to single-drive Macs though. It needs a mirror with a clean copy of the data to correct from.
"Ant are warm blooded, and are going to like heat..."
I don't think that means what you think it means.
I think most people would be okay with ZFS if the license were amenable to inclusion in the Linux kernel, but right now the fact of the matter is that it's not. FUSE isn't acceptable for core functionality, so basically that means Linux doesn't have the option of using ZFS as a successor to any of the major 4 filesystems. If Sun were to change their minds about the license that would quickly change, but Sun picked the license with the specific intent of remaining incompatible with Linux so I would regard that as pretty unlikely.
But, even with Reiser gone I'm not sure the impact on reiserfs3 will be immediate. His company abandoned maintenance on it long ago, I think it's handled by someone in Suse now. It's in a precarious situation anyway, and Reiser would not address that.
Honestly though, I'm not sure the maintenance situation for XFS or JFS is so much better that they solve the problem. If you want an assurance of very good support long term, ext3 is basically your only choice.
BTRFS is the one bright spot on the horizon that I can see. It offers many of the same features as ZFS. It's likely going to be years before it's production ready, but if/when that happens it'll be enough better than the existing Linux filesystems to be the clear successor, and I think it'll be close enough to ZFS that most Linux users will stop pining for ZFS support.
Kudos to Oracle for backing the project.
Three points.
First, yes NIN's making a lot more because of all the attention this has received, but the amount of attention won't be 'zero' if this becomes mainstream. Major internet sites are not the lower bound for word of mouth.
Second, for every big name like NIN that stands to make millions, there's got to be orders of magnitude more people that can make a decent living. I would be very skeptical that this approach goes from millions in a few days to almost nothing for everyone else.
Third, for bands too obscure to make much money doing this, would they really be making a whole lot via traditional CD sales? They'll still have to start out playing seedy bars.
5) Install Firefox 3, which supports real zooming.
Of course, if you're hand-compiling everything, you won't have time to apply a patch every time an exploit is discovered in something you do need. You'll be less secure unless you have implausibly large amounts of time to track and maintain everything.
It's not blindly using vendor binaries, there's nothing blind about it. It's a very explicit cost/benefit analysis. Those vendor binaries save time upfront with the install, they save time with patches and upgrades, they vastly increase the number of servers an admin can maintain, and they make support with whatever vendors you're dealing with much easier.
Indeed.
Most here want neutral ISPs. The only way we're going to get that is if the ISP makes money either way, and their "unlimited" plans aren't the way to do that. They were built around assumptions of usage that are no longer valid with bittorrent, iTunes, YouTube, etc being as big as they are now.
One response is to throttle services that consume too many resources. The other is to bill according to usage and upgrade as necessary. That way, Time Warner has a business case to roll out their upgrades and facilitate higher usage, regardless of what you're doing with it.
A neutral network is metered by the gb.
OTOH, if they see revenue from every gigabyte you download, then they have an incentive to facilitate your usage, rather than transparently throttling it whenever you use a service with an inconveniently large network footprint. As more people download more content, their infrastructure takes a beating with the unlimited plans they now have trouble providing. Per-gigabyte pricing means they have an incentive to upgrade rather than throttle, because there's money in it for them.
Consolidation of ownership notwithstanding, ISPs built their infrastructure and "unlimited" plans around assumptions of usage that are no longer valid. This would be a problem even in a healthy, competitive market. The push to tiered QoS with bandwidth hogs like iTunes and YouTube being throttled is, at least in part, a response to this. ISPs don't want to expose users to the costs of their usage, so they're trying to blackmail it out of everyone else. From a network neutrality perspective, per-gigabyte pricing is great, because it means Time Warner wins whether I'm downloading from iTunes, YouTube, BitTorrrent, etc. It doesn't matter to them because they can simply bill me according to my usage (or, I'll cut back, or, I'll switch to another ISP and take my heavy usage with me).
With pebble bed reactors (and other newer designs), they are "passively safe". This basically means that when the thermal energy in the core gets high enough, it is no longer a neutron-friendly environment and fission can't be sustained. The reactor is designed such that heat is naturally lost to the environment fast enough to keep the equilibrium temperature below that which would melt anything in the core. A meltdown in that case would be impossible, because heat by itself acts like emergency control rods, no operators or computers have to get anything right to halt the reaction.
Erm... it was safe. The reactor suffered a significant failure of its cooling system without escalating beyond that. No design that relies on zero accidents is safe, that requires a design where even accidents aren't catastrophic.
If the sodium leak had involved sodium exposed directly to the reactor (it didn't), environmental contamination would primarily be from Sodium-24, which has a half-life of 15 hours. That's not irreversible contamination, at worst that's a temporary evacuation order. It's not like non-nuclear industrial accidents can't get that bad (eg, the Bhopal disaster). This is something we already have to live with as an industrialized civilization, nuclear power simply has the virtue of releasing far less CO2.
The only part of this that causes me concern is the cover up.
This is merely a money grab!
I certainly hope so.
If they've advertised the service as being "unlimited", they have an incentive transparently throttle people that test the limits of their infrastructure.
If the pricing is based on however many dollars per gigabyte, then they have an incentive to facilitate my usage, not throttle it.
The network in your area may not be that heavily loaded, but that's a big problem in more and more areas, and ISPs are worried about the costs of upgrading their networks. Passing the costs of heavy usage on to the consumer generates the revenue needed to justify upgrades to the network. Or, the user cuts back. Or, they switch to another ISP and take their above average usage with them. I'm fine with all the above.
Figure out what it's going to cost to support my usage, bill me accordingly, and upgrade the infrastructure as necessary. I'll either cut back or bear the cost of my higher usage.
Too tiny to be useful for what? Seems to me it's obviously useful for some stuff, so what you're really saying is that it's not useful for you.
The only problem I've had is with some websites needing horizontal scrolling. I installed Firefox 3 to get zooming and it's fine.
Yes.
Unlike many posters here I actually travel very little. That means I'm not prepared to spend much on a laptop. But, I've found the laptops in my price range tend to be 8+ lb monstrosities. I've got my desktop for heavy lifting, but an Eee is the best choice for very portable and low cost at the same time.
The specs aren't much, but they're enough to get you into Windows XP or a full desktop Linux. You're running the same applications with the peripherals of your choice etc. That's a huge asset.
This tradeoff is not a good one for everyone, but for those of us in a position to benefit, it's the best laptop in years.
I share your skepticism, but it seems likely the new spec will address this. We'll have to wait to see how successful it is.
Indeed.
And remember, you don't need stamps to write your MP!
"The question is how complex is complex enough? The only criteria we know is enough, is interaction with an intelligent observer, a.k.a. observation. Because observation is the only way we can determine the outcome."
Okay. Since the universe is more complex than any of the intelligent observers therein, this shouldn't be a problem by virtue of the fact that the light from the phenomena we're observing would have hit a great many objects without our intervention. We're not dooming the universe with our hubris by catching the light a few nanoseconds before it would have hit the ground anyway.
> For me, it'd be a cost saver.
Sure. Now that virtualization is supported. That's my point. My post refers to the advantages of virtualization.
> It's not clear to me what problem is being solved by having virtual OSX.
The change to their license only applies to OS X server, so it it applies to their server product. I run my domains on a $20/mo VPS service. This is fine for dns, web, e-mail, moderate database, etc. A server like this is more than enough for a small business.
Without support for virtualization, I'd need to buy and colocate an XServe, which would be around an order of magnitude more expensive. This pretty much rules OS X out without even getting into a discussion of the pros and cons, it just can't participate at a price that's interesting.
"As far as I recall, BSD is basically "do what you want as long as the attribution remains""
No. The first clause in a 3 clause BSD license is "Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer, without modification.".
"Nonsense!"
Precisely. Stating bandwidth limits and overage charges upfront is not "unlimited", but sneaking in a bunch of throttling and limitations lets them continue to claim it's unlimited.
"Wouldn't it be simpler for the telcos to charge per GB delivered in addition to the size of the pipe?"
Unacceptable. They wouldn't be able to advertise their service as "unlimited".