Linux has futzed with this a lot (and lets the user tweak VM behavior a lot, but/proc/sys/vm/swappiness goes a long way...), both linux and Windows will swap well ahead of not having free memory (for good reason). Just wanted to go into detail because I keep seeing people complain that they see swap used in linux or windows when they still have free memory, not realizing this isn't a bad thing generally.
There are generally two strategies: -The common-sense one where you swap when you run out of memory. This makes a lot of practical sense on systems with limited write cycles (flash based swap, though you really never ever should do that anyway), and systems that want to spin down drives to conserve power for battery conservation. Performance wise (this may surprise people who haven't spent time thinking about it), this can often be bad. Avoiding swapping is generally only good on systems where resource utilization is carefully managed and you know it won't swap ever (the IO operations of unneeded can interfere with the productive activity of a constantly busy system). This is actually a vast minority of systems in the world (no matter how l33t one may think themselves, they most certainly don't have a usage pattern that would be impacted by the extraneous IO operations of occasional write to swap.
-Pre-emptive swapping. When the IO subsystem is idle and the system can afford to copy memory to swap area, it does so (depending on criteria). Generally speaking it will select memory not accessed much and write it to disk, but leave the memory copy in place if the physical memory is not immediately needed. A fair amount of swap used in an apparently underutilized system is duplicated in physical memory and swap space. The benefit here is that if the process reads back that memory, it doesn't incur any penalty in reading it back despite it being also in swap (the system may make certain decisions on what is the best swap candidate and write to disk different data). The benefit of writing this stuff to swap even when not needed is clear when an application comes along that allocs more memory than the system has free in physical space. In the first strategy, this means the malloc blocks while data is written to disk, and the new application starting or needing a lot of data is severely impacted. In the pre-emptive swap case, system notices the condition, knows what memory it has a backup in swap of that hasn't been used lately, and can free that memory and satisfy the malloc pretty much instantly.
To those who have 1GB of RAM or so it becomes less likely that the system will have to flush memory from physical RAM, but there is a balance to be struck between memory used directly invoked by applications, what the application memory access pattern is, and what ram you can use to buffer filesystem access. If your total application memory allocation is 75%, it still may make sense performance wise to only keep 50% of your physical memory dedicated to the applications, (the other bit relegated to swap), and 50% of the memory to buffer disk I/O.
It's a fair assessment that IBM's open source support is generally targeted at linux. At least on the server side their isn't that much closed though. Standard ATI radeon chip, bnx2 driver NICs (on most of the new stuff, tg3 on x3455), aacraid driver on the SAS end, sata_(chipset name) on the SATA end. The so-called HostRAID they employ on top of SATA I think may still be closed (for the reason anything called HostRAID done in the past has been closed source,) we've just used the basic controller with software RAID on top of it anyway. We don't run any closed source drivers to support the system and go out of the way to avoid them if they come up, and complain loudly if they try to push a closed source driver on us. So, you can support all the useful features of a server with GPL licensed drivers with no closed bits (including things like BMC firmware updates which migrated to OpenIPMI for ease of use). Going forward this shouldn't get worse, they are getting the point that open source drivers are easier to support than closed-source. Their efforts specifically are targetted at linux, and the caveat for BSD developers I suppose is that while the code is available, it is generally GPL licensed which doesn't mesh with BSD. Maybe knowing the linux drivers used gives you a better idea of the BSD support if the newer systems, which I can't honestly speak to.
The remote management I can definitely say on IBM can be done remotely without local presence save for cable connection. Prior to the x3550/x36550, it was the also the case, but with a caveat. The BMC in x336/x346 would always come up initially as 10.1.1.97, so you would have to alias it on some system on the same layer 2 network, and modify it to a sane address by hand remotely (still didn't have to turn on the system, for multiple systems would have to flush arp table between bmc configs, but you get the idea). Systems as of x3550/x3650/x3455/x3755 have BMCs that DHCP by default, which makes it easier to do it fully remotely. Now the one possible issue I might still see is that I generally leverage the DHCP server (add file-name and next-server options) and use PXE to install an OS (or run diskless). What you describe with using the cd in your laptop remotely sounds like you could be referring to the RSA feature of using local media to boot the remote system via the systems management. The RSA does that, but at a price of 300 bucks or so for that card. The IBM built-in BMCs indeed do not implement remote media support. I support larger scale installations and so the remote media approach is unappealing anyway in favor of PXE based installs which are more manageable at large scale, but conversely a little more involved than remote media at extremely small scale (1-2 systems). IBMs strategy for built-in management is that the base user doesn't need remote video, nor a web interface, and will use netboot for remote media needs, and pay a premium if they want those features of 300 dollars. I don't know if you are referring to remote media support or not or if HP's include remote media without paying extra. I know HP used to require the LO card, and don't know if all those features got built-in or if they kept some optional (like IBM integrated a lot of functions into BMCs, but kept RSA alive for 'premium' remote management). I know IBM BMCs are shared on the first NIC, so a lot of times systems I support need at a minimum only one power cable and only one ethernet cable for both the system and the systems management, but RSAs have separate dedicated NIC port. Don't know how HP handles this..
True everyone sells similar stuff nowadays at the commodity level (putting aside HP's itanium, Sun's Ultrasparc, and IBM's power systems, which makes things more complicated), however my experience certainly shows IBM to be capable boxes without need of Windows for everything, with few exceptions. The e325/e326/e326m are out of place and may be subject to your criticism. I don't think of those servers as a sufficiently serious Opteron effort. The x336/x346/ and blades seem pretty good to me, and the IPMI 2.0 based rack mount systems allow SOL in a sane way. The wave of Opteron servers coming are a much more serious effort and work well in general compared to e32*.
My job is exclusively Linux, never ever booting Windows on any of our servers (though admittedly there exists hard drive firmware updates and a few other esoteric updates that are still DOS boot CDs or floppys, however the more common BIOS, BMC, and Diag updates have very good linux support without using DOS at all). In the past they did do goofy things with a powerquest image being written to a linux filesystem with PC-dos and booting into that, and the BMC updates used to require moderately aggravating IBM drivers, but that has been dropped in favor for updates that are self contained (BIOS, diag) or use OpenIPMI drivers (BMC).
All the systems nowadays have similar manageability, ILO is nothing special compared to Dell's, Sun's, and IBM's BMCs nowadays. Everyone sells IPMI compliant management and at least IBM I know implements it well and provides all features I could think of for remote management save for remote video console (but who needs that when you have linux/SOL) without additional cost. RSA cards are there for the people who need remote video console and a fancy web interface. I'd wager everyone's BMC implementation is on par and nowadays manageability is not as much a discriminating factor...
IBM I admit could donate more hardware to some open source efforts, but they do contribute a significant amount of developer work to open source projects, which helps offset the hardware issue some.
Anyway, in summary, IBM may have in the past been subject to that criticism in the x86 space, but in my job experience it has improved greatly.
I wonder how obvious the request for payment was. I could see that if most of the year a bunch of bagels show up, you might look more carefully to determine why they are there and notice the request for payment. I could see that around a holiday season the explanation is 'oh, someone brought in food for the holidays', and grab one without thinking or looking hardly at all. If it was a note in front of the food they might have assumed it probably said something like happy holidays or something, without bothering to read and just grabbing for the food.
It could also serve to explain some of the executive stealing too. I've noticed year round as I talk to executives, they frequently seem to have some sort of food available for people to grab and much on, usually provided or acquired by their administrative assistant. An executive is more likely to be used to random cookies/bagels/muffins/whatever to magically appear for free consumption than us peons at the bottom.
It's people. Hot dog is made out of people. They're making our food out of people. Next thing they'll be breeding us like cattle for food. You've gotta tell them. You've gotta tell them!
Score one for dvorak, my traffic is as secure from the keyboard as the ROT13 I use on the nework... Also for massive amounts of bittorrent traffic in my upstream queue effectively destroying any timing information between my keyboard and the outside world. QoS probably actually works to keep it more intact, but still...
If you were masochistic, have password prompts randomly shift your keymap back and forth while entering your password...
My Windowmaker setup also doesn't have a pager at all, but it does admittedly lack the ability to bring up a view of all the workspaces concurrently (XGL does allow such behavior though).... I will give Apple credit for a relatively innovative backup interface in Time Machine, but Spaces is about as original as Microsoft's Recycle Bin.
http://www.cll.com/articles/article.cfm?articleid= 22 Dr. Seuss v. Penguin Books. "...concluded that the work at issue was not a parody of the Seuss original because it did not ridicule or criticize the prior work, but merely copied the work's best known elements..."
" [T]he substance and content of The Cat In the Hat is not conjured up by the focus on the Brown-Goldman murders or the O.J. Simpson trial. Because there is no effort to create a transformative work with new expression, meaning or message, the infringing work's commercial use further cuts against a fair use defense.15"
Nothing about X inherently gives multiple workspaces/virtual desktops (well, unless you count 'slippery' desktop, but that isn't really considered anything useful), it's just that window managers have for eternity provided it.
Looking at the demo, it only has one functional feature lacking in most common implementations, dynamic reordering of the workspaces. The rest is visual fluff (most of which you can reproduce using XGL). Doesn't do the virtual desktop at all (which most window managers don't seem to do anyway nowadays), where a window could span multiple visible spaces, but rather does the workspace thing of nearly entirely discrete workspaces.
One thing that the video doesn't make clear, including the narration, is whether applications can have windows in multiple workspaces. I.e. when they click an app icon, it switches to 'the workspace where that application is', which I wonder how it decides when the app spans multiple workspaces. The one illustration of a window being moved was a single-window application. It doesn't show whether or not an entire application would move with it. I remember one early OSX implementation did workspaces by twiddling with application hiding, which induced such a restriction. I can't imagine Apple would end up with the restriction, but the video didn't do a good job explaining the degree to which their implementation is flexible.
All in all an important step to a serious power user desktop out of the box, but nothing significant to say it's different from traditional unix window manager implementations. It's not a bad thing (I can't think of much of a way to improve on the model), but just because Apple does it doesn't mean it's magically better than everyone else's.
If we are playing the whole intellectual property game (which we conveniently like to do when someone we *don't* like pulls this stuff), did they: -Get the permission of DC to use the likeness of 'The Penguin' in making over Al Gore? -Get the permission of Marvel for using X-Men 3 imagery?
So they managed to rip off the Linux logo, and both of the major comic publishers, they really wanted to piss geeks off...
Copyrighted material may be reproduced in parody if it is the subject of the parody. However, legally, if the copyrighted material is being used to parody something else, then it's not technically legal.
For example, Penny Arcade had to remove a comic featuring Strawberry Shortcake because it infringed on American Greeting's copyright. Penny Arcade was using the character to parody American Mcgee's style of game development.
Maybe GNUstep and its ilk would get more development attention if OSX were king (along the lines of the attention wine gets today).
In any event, Apple's got a business model they achieve relative success with and know how to manage it. They don't know how transitioning to be of a software company would pay off or if they would be able to deal with the transition easier. The software is largely subsidized by their hardware sales today, with the standalone purchase revenue of OSX probably being little more than token cash flow.
OSX is, business-wise, largely a hype generator for the Mac hardware. If it weren't exclusive, that would go away. More likely than them making OSX a wider product would be ditching OSX as everyone knows it today and instead doing a Windows based platform, but adding the 'hype-worthy' features of OSX. Also not likely I at least hope, but I would say more likely than Apple trying to become more of a software vendor.
Already the article suggests it may not be capable of running linpack, the other question being, are these 32-bit precision operations or 64-bit precision? Linpack explicitly measures 64-bit precsion. This is one reason why despite some clustered deployments that are inevitible with the cell processor, those won't be impressive top500 wise despite the cries of 'OMFG, cell has uber gigaflops'. Cell brags on the gigaflops, but the state of Cell as it is announced today is only interesting 32-bit precision wise. 64-bit precision won't blow away the conventional Power/PPC chips which are impressive Linpack wise.
Roosevelt served 4 terms, and because of him we have the two term limit. It would be a 'first since FDR', but it wouldn't be an absolute first. It would mean more because he'd have to violate the two term legislated limit, which FDR didn't have to do.
The closest equivalent to 'buffered' would probably be registered. Only FB-DIMMs will be flagged as 'fully buffered' because, well, FB-DIMM stands for Fully Buffered DIMM.
Yet another technology accepting higher latency in favor of higher throughput (like Rambus, but less restricted). DDR2 is higher latency than DDR1.
Motherboards may become cheaper, as the traces for memory are much simpler for FB-DIMMs than traditional DIMMs.
You could have massive amounts of horribly slow memory and have high Nmax. To some extent large N will generally show the performance of a cluster better..
BlueGene processors are very slow, but compensate by having an incredible architecture for inter-processor communication, enabling BlueGene to scale very well. Each BlueGene processor is less than a GHz, and while power based systems are generally more clock-efficient than others, a single BlueGene processor would look puny next to even mediocre desktop chips these days, whereas most clusters are comprised of processors that would compare favorably with a typical desktop each.
Rmax represents the maximum acheived measured FLOPs as a result of an xhpl run.
Nmax represents the problem size. Nmax generally is aimed to be a problem that consumes as much memory as possible without swapping.
Rpeak is the theoretical max FLOPs possible according to the processor used. For example, a PPC chip is theoretically capable of 4 Flops per clock, so multiply the clock by the number of cores in the cluster. x86_64 is theoretically capable of 2 flops per clock, so multiply cores by two. Note that AMD clock for clock doesn't do any better than intel in *this* particular benchmark, so Intel clusters inherently can climb this list better, despite poor memory performance and other factors that make them less useful in a general supercomputing sense. Itanium can acheive better floating point (I believe 8 flops per clock).
And for anyone seeking to compare Rpeak/Rmax numbers with published Cell figures, keep in mind that game consoles (and by extension cell) brag about their single precision (32-bit) floating point performance, whereas this list only deals with double precision numbers (64 bit). Cell actually is nothing special at get top500 relevant benchmark results.
Many people feel this very specific benchmark is a poor indicator of the overall effectiveness of a cluster, and consider hpcc (which includes hpl as a subset) to be a better holistic method to evaluate the value of a cluster.
I know by itself thermite and similar methods have difficulty penetrating the outer case reliably, but I would think drill+thermite injection to fill the internal cavity of the system would be effective..
Combined with an encryption scheme I would think it virtually impossilbe to recover data if you can reduce the platters to slag reliably..
The problem I have is that the cargo area is not pressurized (why you need to be very cognizant of bottles of stuff), so it could be very lethal to attempt such a stunt.
I seriously doubt they take real issue with long, plot-driven rpg/adventure games. Obviously a core aspect of such games is having a world to explore and part of that is the 'tedium' of getting what is needed to progress.
What I do understand them getting aggravated about are games that are immediate action games with unlockable features. The shining example here is a fighting game where stages, costumes, characters are largely limited at the beginning and unlock over usage. I understand the cosmetic aspects of costumes and stages in such games being unlockable, but fundamentally interesting characters with different moves and such are also precluded. This is very aggravating because these games tend to be group games and it isn't so much fun when people come together to see a bunch of '?' characters.
The primary incentive for the game developers I can conceive is to deter rentals. It is not so easy to go and rent a fighting game as a one off deal for a get together and have the full experience with these impediments in place.
Slashdot Mirror
Linux has futzed with this a lot (and lets the user tweak VM behavior a lot, but /proc/sys/vm/swappiness goes a long way...), both linux and Windows will swap well ahead of not having free memory (for good reason). Just wanted to go into detail because I keep seeing people complain that they see swap used in linux or windows when they still have free memory, not realizing this isn't a bad thing generally.
There are generally two strategies:
-The common-sense one where you swap when you run out of memory. This makes a lot of practical sense on systems with limited write cycles (flash based swap, though you really never ever should do that anyway), and systems that want to spin down drives to conserve power for battery conservation. Performance wise (this may surprise people who haven't spent time thinking about it), this can often be bad. Avoiding swapping is generally only good on systems where resource utilization is carefully managed and you know it won't swap ever (the IO operations of unneeded can interfere with the productive activity of a constantly busy system). This is actually a vast minority of systems in the world (no matter how l33t one may think themselves, they most certainly don't have a usage pattern that would be impacted by the extraneous IO operations of occasional write to swap.
-Pre-emptive swapping. When the IO subsystem is idle and the system can afford to copy memory to swap area, it does so (depending on criteria). Generally speaking it will select memory not accessed much and write it to disk, but leave the memory copy in place if the physical memory is not immediately needed. A fair amount of swap used in an apparently underutilized system is duplicated in physical memory and swap space. The benefit here is that if the process reads back that memory, it doesn't incur any penalty in reading it back despite it being also in swap (the system may make certain decisions on what is the best swap candidate and write to disk different data). The benefit of writing this stuff to swap even when not needed is clear when an application comes along that allocs more memory than the system has free in physical space. In the first strategy, this means the malloc blocks while data is written to disk, and the new application starting or needing a lot of data is severely impacted. In the pre-emptive swap case, system notices the condition, knows what memory it has a backup in swap of that hasn't been used lately, and can free that memory and satisfy the malloc pretty much instantly.
To those who have 1GB of RAM or so it becomes less likely that the system will have to flush memory from physical RAM, but there is a balance to be struck between memory used directly invoked by applications, what the application memory access pattern is, and what ram you can use to buffer filesystem access. If your total application memory allocation is 75%, it still may make sense performance wise to only keep 50% of your physical memory dedicated to the applications, (the other bit relegated to swap), and 50% of the memory to buffer disk I/O.
didn't make a huge difference in the quality of lightning,
You had some high expectations there didn't you?
It's a fair assessment that IBM's open source support is generally targeted at linux. At least on the server side their isn't that much closed though. Standard ATI radeon chip, bnx2 driver NICs (on most of the new stuff, tg3 on x3455), aacraid driver on the SAS end, sata_(chipset name) on the SATA end. The so-called HostRAID they employ on top of SATA I think may still be closed (for the reason anything called HostRAID done in the past has been closed source,) we've just used the basic controller with software RAID on top of it anyway. We don't run any closed source drivers to support the system and go out of the way to avoid them if they come up, and complain loudly if they try to push a closed source driver on us. So, you can support all the useful features of a server with GPL licensed drivers with no closed bits (including things like BMC firmware updates which migrated to OpenIPMI for ease of use). Going forward this shouldn't get worse, they are getting the point that open source drivers are easier to support than closed-source. Their efforts specifically are targetted at linux, and the caveat for BSD developers I suppose is that while the code is available, it is generally GPL licensed which doesn't mesh with BSD. Maybe knowing the linux drivers used gives you a better idea of the BSD support if the newer systems, which I can't honestly speak to.
The remote management I can definitely say on IBM can be done remotely without local presence save for cable connection. Prior to the x3550/x36550, it was the also the case, but with a caveat. The BMC in x336/x346 would always come up initially as 10.1.1.97, so you would have to alias it on some system on the same layer 2 network, and modify it to a sane address by hand remotely (still didn't have to turn on the system, for multiple systems would have to flush arp table between bmc configs, but you get the idea). Systems as of x3550/x3650/x3455/x3755 have BMCs that DHCP by default, which makes it easier to do it fully remotely. Now the one possible issue I might still see is that I generally leverage the DHCP server (add file-name and next-server options) and use PXE to install an OS (or run diskless). What you describe with using the cd in your laptop remotely sounds like you could be referring to the RSA feature of using local media to boot the remote system via the systems management. The RSA does that, but at a price of 300 bucks or so for that card. The IBM built-in BMCs indeed do not implement remote media support. I support larger scale installations and so the remote media approach is unappealing anyway in favor of PXE based installs which are more manageable at large scale, but conversely a little more involved than remote media at extremely small scale (1-2 systems). IBMs strategy for built-in management is that the base user doesn't need remote video, nor a web interface, and will use netboot for remote media needs, and pay a premium if they want those features of 300 dollars. I don't know if you are referring to remote media support or not or if HP's include remote media without paying extra. I know HP used to require the LO card, and don't know if all those features got built-in or if they kept some optional (like IBM integrated a lot of functions into BMCs, but kept RSA alive for 'premium' remote management). I know IBM BMCs are shared on the first NIC, so a lot of times systems I support need at a minimum only one power cable and only one ethernet cable for both the system and the systems management, but RSAs have separate dedicated NIC port. Don't know how HP handles this..
Don't know about Dell or Sun honestly.
Disclaimer: I happen to be an IBM employee
True everyone sells similar stuff nowadays at the commodity level (putting aside HP's itanium, Sun's Ultrasparc, and IBM's power systems, which makes things more complicated), however my experience certainly shows IBM to be capable boxes without need of Windows for everything, with few exceptions. The e325/e326/e326m are out of place and may be subject to your criticism. I don't think of those servers as a sufficiently serious Opteron effort. The x336/x346/ and blades seem pretty good to me, and the IPMI 2.0 based rack mount systems allow SOL in a sane way. The wave of Opteron servers coming are a much more serious effort and work well in general compared to e32*.
My job is exclusively Linux, never ever booting Windows on any of our servers (though admittedly there exists hard drive firmware updates and a few other esoteric updates that are still DOS boot CDs or floppys, however the more common BIOS, BMC, and Diag updates have very good linux support without using DOS at all). In the past they did do goofy things with a powerquest image being written to a linux filesystem with PC-dos and booting into that, and the BMC updates used to require moderately aggravating IBM drivers, but that has been dropped in favor for updates that are self contained (BIOS, diag) or use OpenIPMI drivers (BMC).
All the systems nowadays have similar manageability, ILO is nothing special compared to Dell's, Sun's, and IBM's BMCs nowadays. Everyone sells IPMI compliant management and at least IBM I know implements it well and provides all features I could think of for remote management save for remote video console (but who needs that when you have linux/SOL) without additional cost. RSA cards are there for the people who need remote video console and a fancy web interface. I'd wager everyone's BMC implementation is on par and nowadays manageability is not as much a discriminating factor...
IBM I admit could donate more hardware to some open source efforts, but they do contribute a significant amount of developer work to open source projects, which helps offset the hardware issue some.
Anyway, in summary, IBM may have in the past been subject to that criticism in the x86 space, but in my job experience it has improved greatly.
I wonder how obvious the request for payment was. I could see that if most of the year a bunch of bagels show up, you might look more carefully to determine why they are there and notice the request for payment. I could see that around a holiday season the explanation is 'oh, someone brought in food for the holidays', and grab one without thinking or looking hardly at all. If it was a note in front of the food they might have assumed it probably said something like happy holidays or something, without bothering to read and just grabbing for the food.
It could also serve to explain some of the executive stealing too. I've noticed year round as I talk to executives, they frequently seem to have some sort of food available for people to grab and much on, usually provided or acquired by their administrative assistant. An executive is more likely to be used to random cookies/bagels/muffins/whatever to magically appear for free consumption than us peons at the bottom.
Just putting forth an alternative explanation.
I get my HD content with rabbit ears.
It's people. Hot dog is made out of people. They're making our food out of people. Next thing they'll be breeding us like cattle for food. You've gotta tell them. You've gotta tell them!
No no no, they did Take On Me, these guys are one of those other groups from back then.
Score one for dvorak, my traffic is as secure from the keyboard as the ROT13 I use on the nework... Also for massive amounts of bittorrent traffic in my upstream queue effectively destroying any timing information between my keyboard and the outside world. QoS probably actually works to keep it more intact, but still...
If you were masochistic, have password prompts randomly shift your keymap back and forth while entering your password...
My Windowmaker setup also doesn't have a pager at all, but it does admittedly lack the ability to bring up a view of all the workspaces concurrently (XGL does allow such behavior though).... I will give Apple credit for a relatively innovative backup interface in Time Machine, but Spaces is about as original as Microsoft's Recycle Bin.
http://www.cll.com/articles/article.cfm?articleid= 22
Dr. Seuss v. Penguin Books.
"...concluded that the work at issue was not a parody of the Seuss original because it did not ridicule or criticize the prior work, but merely copied the work's best known elements..."
" [T]he substance and content of The Cat In the Hat is not conjured up by the focus on the Brown-Goldman murders or the O.J. Simpson trial. Because there is no effort to create a transformative work with new expression, meaning or message, the infringing work's commercial use further cuts against a fair use defense.15"
Nothing about X inherently gives multiple workspaces/virtual desktops (well, unless you count 'slippery' desktop, but that isn't really considered anything useful), it's just that window managers have for eternity provided it.
Looking at the demo, it only has one functional feature lacking in most common implementations, dynamic reordering of the workspaces. The rest is visual fluff (most of which you can reproduce using XGL). Doesn't do the virtual desktop at all (which most window managers don't seem to do anyway nowadays), where a window could span multiple visible spaces, but rather does the workspace thing of nearly entirely discrete workspaces.
One thing that the video doesn't make clear, including the narration, is whether applications can have windows in multiple workspaces. I.e. when they click an app icon, it switches to 'the workspace where that application is', which I wonder how it decides when the app spans multiple workspaces. The one illustration of a window being moved was a single-window application. It doesn't show whether or not an entire application would move with it. I remember one early OSX implementation did workspaces by twiddling with application hiding, which induced such a restriction. I can't imagine Apple would end up with the restriction, but the video didn't do a good job explaining the degree to which their implementation is flexible.
All in all an important step to a serious power user desktop out of the box, but nothing significant to say it's different from traditional unix window manager implementations. It's not a bad thing (I can't think of much of a way to improve on the model), but just because Apple does it doesn't mean it's magically better than everyone else's.
If we are playing the whole intellectual property game (which we conveniently like to do when someone we *don't* like pulls this stuff), did they:
-Get the permission of DC to use the likeness of 'The Penguin' in making over Al Gore?
-Get the permission of Marvel for using X-Men 3 imagery?
So they managed to rip off the Linux logo, and both of the major comic publishers, they really wanted to piss geeks off...
Copyrighted material may be reproduced in parody if it is the subject of the parody. However, legally, if the copyrighted material is being used to parody something else, then it's not technically legal.
For example, Penny Arcade had to remove a comic featuring Strawberry Shortcake because it infringed on American Greeting's copyright. Penny Arcade was using the character to parody American Mcgee's style of game development.
Maybe GNUstep and its ilk would get more development attention if OSX were king (along the lines of the attention wine gets today).
In any event, Apple's got a business model they achieve relative success with and know how to manage it. They don't know how transitioning to be of a software company would pay off or if they would be able to deal with the transition easier. The software is largely subsidized by their hardware sales today, with the standalone purchase revenue of OSX probably being little more than token cash flow.
OSX is, business-wise, largely a hype generator for the Mac hardware. If it weren't exclusive, that would go away. More likely than them making OSX a wider product would be ditching OSX as everyone knows it today and instead doing a Windows based platform, but adding the 'hype-worthy' features of OSX. Also not likely I at least hope, but I would say more likely than Apple trying to become more of a software vendor.
Already the article suggests it may not be capable of running linpack, the other question being, are these 32-bit precision operations or 64-bit precision? Linpack explicitly measures 64-bit precsion. This is one reason why despite some clustered deployments that are inevitible with the cell processor, those won't be impressive top500 wise despite the cries of 'OMFG, cell has uber gigaflops'. Cell brags on the gigaflops, but the state of Cell as it is announced today is only interesting 32-bit precision wise. 64-bit precision won't blow away the conventional Power/PPC chips which are impressive Linpack wise.
Roosevelt served 4 terms, and because of him we have the two term limit. It would be a 'first since FDR', but it wouldn't be an absolute first. It would mean more because he'd have to violate the two term legislated limit, which FDR didn't have to do.
The closest equivalent to 'buffered' would probably be registered. Only FB-DIMMs will be flagged as 'fully buffered' because, well, FB-DIMM stands for Fully Buffered DIMM.
Yet another technology accepting higher latency in favor of higher throughput (like Rambus, but less restricted). DDR2 is higher latency than DDR1.
Motherboards may become cheaper, as the traces for memory are much simpler for FB-DIMMs than traditional DIMMs.
You could have massive amounts of horribly slow memory and have high Nmax. To some extent large N will generally show the performance of a cluster better..
BlueGene processors are very slow, but compensate by having an incredible architecture for inter-processor communication, enabling BlueGene to scale very well. Each BlueGene processor is less than a GHz, and while power based systems are generally more clock-efficient than others, a single BlueGene processor would look puny next to even mediocre desktop chips these days, whereas most clusters are comprised of processors that would compare favorably with a typical desktop each.
Rmax represents the maximum acheived measured FLOPs as a result of an xhpl run.
Nmax represents the problem size. Nmax generally is aimed to be a problem that consumes as much memory as possible without swapping.
Rpeak is the theoretical max FLOPs possible according to the processor used. For example, a PPC chip is theoretically capable of 4 Flops per clock, so multiply the clock by the number of cores in the cluster. x86_64 is theoretically capable of 2 flops per clock, so multiply cores by two. Note that AMD clock for clock doesn't do any better than intel in *this* particular benchmark, so Intel clusters inherently can climb this list better, despite poor memory performance and other factors that make them less useful in a general supercomputing sense. Itanium can acheive better floating point (I believe 8 flops per clock).
And for anyone seeking to compare Rpeak/Rmax numbers with published Cell figures, keep in mind that game consoles (and by extension cell) brag about their single precision (32-bit) floating point performance, whereas this list only deals with double precision numbers (64 bit). Cell actually is nothing special at get top500 relevant benchmark results.
Many people feel this very specific benchmark is a poor indicator of the overall effectiveness of a cluster, and consider hpcc (which includes hpl as a subset) to be a better holistic method to evaluate the value of a cluster.
I know by itself thermite and similar methods have difficulty penetrating the outer case reliably, but I would think drill+thermite injection to fill the internal cavity of the system would be effective..
Combined with an encryption scheme I would think it virtually impossilbe to recover data if you can reduce the platters to slag reliably..
Well, advertising yeast to a question, that is a tough one.
Link to story?
The problem I have is that the cargo area is not pressurized (why you need to be very cognizant of bottles of stuff), so it could be very lethal to attempt such a stunt.
I seriously doubt they take real issue with long, plot-driven rpg/adventure games. Obviously a core aspect of such games is having a world to explore and part of that is the 'tedium' of getting what is needed to progress.
What I do understand them getting aggravated about are games that are immediate action games with unlockable features. The shining example here is a fighting game where stages, costumes, characters are largely limited at the beginning and unlock over usage. I understand the cosmetic aspects of costumes and stages in such games being unlockable, but fundamentally interesting characters with different moves and such are also precluded. This is very aggravating because these games tend to be group games and it isn't so much fun when people come together to see a bunch of '?' characters.
The primary incentive for the game developers I can conceive is to deter rentals. It is not so easy to go and rent a fighting game as a one off deal for a get together and have the full experience with these impediments in place.
For PC market true, but for a good linux laptop, thinkpads have of late been an excellent choice with all the features working under linux.