That sounds good, but there would be a flamewar and debate for every bugfix release. We already have a better way, we define increments in the minor number when new major subsystems or core improvements are added en-masse.
severe problems with consoles on the Radeon framebuffer device
I did too, but once I switched to VESA framebuffer, everything was fixed, much less of a headache to use VESA unless you have the need for hardware-accelerated framebuffer. I just use FB for the console, so hardware acceleration seeemed like it wasn't worth the headache.
As for CD burning, mine's much better since 2.6 came out. I particularly like that I no longer have to emulate SCSI over IDE to get my burner up-and-running. Have you tried another brand burner, just for kicks? I've had several devices/configs work in Windows, but not Linux, and vice-versa.
What are you doing on the framebuffer that requires the accelerated driver, and what ancient burning software are you using that doesn't handle 2.6 packet-writing?
What exactly are you doing that 2.6 isn't cooked-enough for your needs yet?
I'm really curious because I felt that after the disaster that the early 2.4 series was, the kernel team really pushed a good 2.6 release out and it's been quite smooth from 2.6.5.
Are you running strange hardware or binary-only drivers or something?
Because bumps to the major version number indicate HUGE-scale rewrites, while the minor (.6 in this case) define feature-complete stable branches, and the trailing number at the end is for bugfixes and minor enhancements.
This is the way software SHOULD be versioned. It's the way Apple is versioning now, and it's the way Microsoft versions it's core systems (Windows XP SP2 = NT 5.1.2600).
Personally, I'd like for the odd-minor devel releases to go away and find some better way of versioning those, but everything else to-date has been sensible and sane, and I've been compiling my own kernels since the 2.1 series.
Linux gets faster and faster too, the improvements in speed on OS X come mostly from Apple's contributions to GCC for the PowerPC architecture. I built a gentoo system with GCC-2.95 and then rebuilt with GCC-3.4 and there was a significant performance boost.
Other improvements come from code cleanups and kernel tweaks, which the Linux side has seen quite a bit of with the recent XOrg and KDE releases.
I admin the Macs, which are only 200 of over 2000 machines. The servers are all W2K. I'd like to run OpenLDAP+SAMBA on the x86 servers we already have, and transition away from AD.
The question is really if there are any Linux distros that can plug into AD or OD networks, as clients or servers, without having to fiddle too much.
Thats what I meant too, though to make the question quick I just asked mostly about client-end stuff.
Apple's OS X doesn't use/etc/passwd files by default, it uses a database called NetInfo, and it's pretty cool. I'm very curious as to why there aren't any Linux distros taking things another step and using OpenLDAP to handle accounts from the first time you boot it.
Because 'the people upstairs' who make purchasing decisions are dead-set on x86 hardware in the server room. Also, there's perfectly good x86 hardware in there now, I'd rather use itr than pay Apple for new metal.
I'm a bit upset, it seems like there oughtta be a distro based on LDAP out there, a distro without the default user and group info in/etc/passwd, but stored in LDAP. This same distro oughtta ship with as many services kerberized as possible, and a pre-configured KDC or setup tool to link to an existing one.
The reason I'm asking is because I've been using the OS X directory services, and just got 200 Macs onto the AD, and it's a beautiful thing. It's much more convoluted to do the same in Linux, and one would think that there would be some sort of similar tool to handle directory-service kung-fu.
I'm just concerned that Linux will have a lot of trouble getting into the mid-sized and small shops because it doesn't interoperate well out-of-the-box, to connect a Linux box to an AD is a total pain in the arse, serving OpenLDAP is even more of a pain.
The space is automatically placed by slashcode to prevent robotic shenanigans, I believe.
If you look back, open video drivers have ALWAYS lagged months or years behind the hardware releases, it takes a long time for the developers to figure out how to access the new functions of the cards. I've been using XFree86 since the 3.2 days and I always knew that you could never run it on cutting-edge hardware. Hell, it used to be worse than not having 3D, a lot of new cards back then PLAIN WOULDN'T WORK.
As for what I linked to, those are the open drivers included with xorg, they are the drivers produced by and for the x.org community. ATI's drivers are something else altogether, and support 3D in the newer cards.
You still can do 3D on the "2D only" cards, it'll just be dirt slow because it's emulated. You'll get better 3D performance from an ancient Rage 128Pro than an X800, but the 9200 will give you the best open-source 3D experience right now.
A Randolph, Massachusetts resident going by the handle 'MarcQuadra' has discovered a briefcase that magically lowers his heating bill by 90%. When asked about it he replied only that he can now afford an LCD screen, but his eye fluids have all boiled off.
the streets in the Middle East would flow with blood to an unprecedented degree.
This is another popular misconception. Israel has one of the most powerful millitaries in the world, and they are a nuclear power; an invading army would be destroyed quite handily by IDF, and Israel wouldn't hesitate to use a small nuke on the invader's capitol. For this reason, Israel is safe from invasion. Individuals might be suicidal, but nations are not.
As for the Palestinians already inside 'greater Israel', the reason they fight with suicide bombs and crude mortars is because they have nothing else to fight with, they pose little to no threat to Israel on the whole. And also, our aid doesn't change their situation, it only aligns us with their oppressors.
We sow discontent across the world for standing unquestioningly with a fundamentalist government that feels no remorse bulldozing an entire block of houses because one has a tunnel used for smuggling arms in it. Tell me that we would accept that sort of behavior from anyone else and still call them an ally.
Our entire relationship with Israel was born out of cold war paranoia that the middle-east would go communist and a powerful domestic Jewish lobby. The population of America stands for it to this day, when the relationship serves us no purpose but to make us the target of terror, because of an irrational and media-induced American guilt for the holocaust.
I've had this argument with several of my Jewish friends, several of whom are stalwart neo-zionists. There would be no terror to wage war on if we didn't so blatantly ask for it.
As for other countries in the middle east, they're all U.N signatories, and just because we have a really big fscking millitary doesn't obligate us to make sure the maps don't have to be redrawn. Our influence on the oil producing middle east countries is arguably the leading reason those people haven't stood up and wrought governments that suit them better, our foreign policy to-date has been to line the pocket of whatever dictatioial regime will sell oil to us.
As a superpower, we should be aware that hubris will most likely be our downfall. Our foreign policy should be as hands-of as possible, let the U.N. come to decisions involving territory and soverignty, thats why it was created. If we follow our current path, we'll come head-to-head with China eventually, when they realize that we've got interests surrounding them and they stop loaning us money.
But Bin Laden doesn't WANT unorganized chaos and death. This is a common misconception of mose Americans. If you actually READ the stuff Bin Laden says, his goal is to get us and our influence out of Muslim lands.
If Bin Laden wanted to kill as many Americans as possible, there'd be people getting shot at malls and suicide bombs in America EVERY DAY. Trust me, there's a LOT of available suicide manpower here in the U.S., they just aren't tapped beause the goal of terror is to make a point and get your needs met.
We could stop AlQaeda in ONE DAY if we stopped giving Israel (a leading EXPORTER of arms) aid and a blind eye, and brought our 'stabilization troops' that prop-up the House of Saud back home. Instead we march right into the foray at great human, moral, and financial cost.
If AlQaeda made a computer virus, it would have a payload that showed messages on the screen of your machine like "Stop supporting Israel and I'll stop planning attacks." or "There were NO beheadings before Abu-Gharib, and we behead only those directly involved with the occupation" or even "Click -HERE- to see what U.S./Israeli millitary action inspired me to take revenge on YOUR towers."
I can EMULATE x86 on a PowerPC 750 and get better performance/watt than VIA's EDEN chips. Really, VIA chips have always been seriously weak sauce, just because they can't clock them well and they use not-much juice doesn't make them superior in any way.
The Pentium-M and AMD GEODE NX CPUs produce MUCH more horsepower per watt than VIA's chips. VIA's only advantage is that they mass-produce Mini-ITX boards and sell them to distributors. If someone mass-produced a GEODE NX bMini-ITX board, it would wipe the floor with the EPIA, fanlessly.
I just bought video cards a few weeks ago. the RADEON 9200 series is fully-accelerated with open-source drivers, and it doesn't need a fan, which is awesome because my system is whisper-quiet now.
It's also plenty fast for what I do with it, I'm only 2D for 99% of the time I'm on the machine anyway, and when I do play 3D games they're a year or two old, so they run fine.
I'm glad we're finding truth here, though I'd like to point out one more thing, I have a 'last laugh' personality.:-)
My point was that it's simply easier and more efficient to pass those offsets around in single registers and do calculations in single operations.
True, but how much LESS efficient is the other overhead of the OS when running full 64-bit? If large disk access incurs 5% less overhead, but the OS uses 20% more memory to run itself and apps, are we really gaining anything? (numbers shamelessly pulled from my arse)
I think the innovation (bus speed boosts, hypertransport, PCI-express, etc) will be happening on the 64-bit platforms now, so the reason to 'go 64' will be the better features and forward compatability, not the bitness of the system itself.
I wonder when Linux users will stop buying hardware that doesn't have published interfaces.
I also wonder when people with ATI card that are actually supported will realize it. My RADEON 9200 and 7500 get full 3D acceleration without the closed drivers.
Having lots of ram in a fileserver can help with caching/buffering of data from disk.
yes, but returns diminish after you exceed your 80/20 'working set' which at any given moment is usually under 2GB.
files of 2G or more are not that uncommon. DVD's and raw audio/video can easily produce multi-gigabyte files. For 32-bit cpu's to work on large files, various kludges and work-arounds have to be used.
Not really, 32-bit CPUs easily handle file I/O operations in the multi-gigabyte range, it's 64-bit precision calculations that 64-bit CPUs benefit from. Even if there is some overhead from dealing with huge files, the VAST majority of a file server's performance issues stem from disk and network subsystem performance. My file server regularly handles files over 4GB (DV files) and the CPU level rarely exceeds 10%, and that's on a 350MHz machine.
Before then, such files could only be made and accessed on 64-bit platforms.
IIRC, the file handling for even the 64-bit flavors of Linux were 32-bit before the mid 2.4 series. Both platforms got this at the same time.
And hard disks (let alone RAID arrays[sic]) are getting so large that the sectors can't be addressed with 32-bit integers (4G * 512 = 2T).
Sorry, that's not making any sense to me. ATA subsystems have been using 48-bit LBA addresses for a while now, and SCSI uses similarly huge numbers. You assume that each byte of the drive needs to be addressed, but this is not true, each SECTOR needs to be addressed, so you don't have to multiply by 512B/sector. Drives speak only in sectors to their controllers, it's the software component of the filesystem driver that decides what the contents of the sector will be.
Like I said, 64-bit file servers don't really have any advantage over 32-bit ones today, unless you need more than 2GB RAM. They WILL become popular before their time though, because of misconceptions like yours and the falling price of 64-bit computing in the marketplace.
I'd also like to note that I'm not against 64-bit computing, I'd rather move up before we hit the limits of today's hardware. But I think we have to be educated about what we can alreay do with 32-bit computers. I estimate that I'll be peachy-keen with my own 32-bit machines for about three more years. I'm suggesting 64-bit file servers at work now, actually, because I know that AMD's Opterons are overall better, more efficient, and cheaper than Intel's server chips; not to mention that it'll be the best way to finally get rid of our Dell servers (pray for me).
Because the mini doesn't need ANY fan at all. The mini cools by convection it it's tiny plastic shell, it'll be even cooler in a PC case.
If you place stuff correctly in a PC case you can even get the heated air from the CPU to draw in cool air over the hard drive.
Because the mini runs OS X and Linux, the Opteron runs Windows and Linux. OS X is an awesome no-hassle server platform.
Because 64-bit doesn't mean ANYTHING when you're a file server with under 4GB of RAM. I'd rather NOT have a 64-bit computer right now, I'm quite happy with 32 until there's a compelling reason to move up. File servers have no reason whatsoever to go 64-bit unless 64-bit parts becomes cheaper through volume than 32-bit.
I'm considering doing this because my 300MHz G3 tower is getting a bit long in the tooth, and I'm doing more and more with it these days. Stepping up to a 1.25GHz G4 would be quite nice, but I won't do it unless I can put two hard drives on at once.
Re:Gave up because the installer wouldn't let them
on
iPod Shuffle RAID
·
· Score: 1
Ahh. You CAN boot off a software RAID in OS X, but it has to be on a supported bus. The USB and FireWire busses do not have RAID drivers in firmware, IIRC. Anyone who wants to do RAID on firewire or USB is foolish anyway, the bottleneck is the bus and the 'glue' chip, not the disk.
Slashdot Mirror
That sounds good, but there would be a flamewar and debate for every bugfix release. We already have a better way, we define increments in the minor number when new major subsystems or core improvements are added en-masse.
severe problems with consoles on the Radeon framebuffer device
I did too, but once I switched to VESA framebuffer, everything was fixed, much less of a headache to use VESA unless you have the need for hardware-accelerated framebuffer. I just use FB for the console, so hardware acceleration seeemed like it wasn't worth the headache.
As for CD burning, mine's much better since 2.6 came out. I particularly like that I no longer have to emulate SCSI over IDE to get my burner up-and-running. Have you tried another brand burner, just for kicks? I've had several devices/configs work in Windows, but not Linux, and vice-versa.
What are you doing on the framebuffer that requires the accelerated driver, and what ancient burning software are you using that doesn't handle 2.6 packet-writing?
What exactly are you doing that 2.6 isn't cooked-enough for your needs yet?
I'm really curious because I felt that after the disaster that the early 2.4 series was, the kernel team really pushed a good 2.6 release out and it's been quite smooth from 2.6.5.
Are you running strange hardware or binary-only drivers or something?
Because bumps to the major version number indicate HUGE-scale rewrites, while the minor (.6 in this case) define feature-complete stable branches, and the trailing number at the end is for bugfixes and minor enhancements.
This is the way software SHOULD be versioned. It's the way Apple is versioning now, and it's the way Microsoft versions it's core systems (Windows XP SP2 = NT 5.1.2600).
Personally, I'd like for the odd-minor devel releases to go away and find some better way of versioning those, but everything else to-date has been sensible and sane, and I've been compiling my own kernels since the 2.1 series.
You first.
Linux gets faster and faster too, the improvements in speed on OS X come mostly from Apple's contributions to GCC for the PowerPC architecture. I built a gentoo system with GCC-2.95 and then rebuilt with GCC-3.4 and there was a significant performance boost.
Other improvements come from code cleanups and kernel tweaks, which the Linux side has seen quite a bit of with the recent XOrg and KDE releases.
I admin the Macs, which are only 200 of over 2000 machines. The servers are all W2K. I'd like to run OpenLDAP+SAMBA on the x86 servers we already have, and transition away from AD.
The question is really if there are any Linux distros that can plug into AD or OD networks, as clients or servers, without having to fiddle too much.
Thats what I meant too, though to make the question quick I just asked mostly about client-end stuff.
/etc/passwd files by default, it uses a database called NetInfo, and it's pretty cool. I'm very curious as to why there aren't any Linux distros taking things another step and using OpenLDAP to handle accounts from the first time you boot it.
Apple's OS X doesn't use
Because 'the people upstairs' who make purchasing decisions are dead-set on x86 hardware in the server room. Also, there's perfectly good x86 hardware in there now, I'd rather use itr than pay Apple for new metal.
I'm a bit upset, it seems like there oughtta be a distro based on LDAP out there, a distro without the default user and group info in /etc/passwd, but stored in LDAP. This same distro oughtta ship with as many services kerberized as possible, and a pre-configured KDC or setup tool to link to an existing one.
The reason I'm asking is because I've been using the OS X directory services, and just got 200 Macs onto the AD, and it's a beautiful thing. It's much more convoluted to do the same in Linux, and one would think that there would be some sort of similar tool to handle directory-service kung-fu.
I'm just concerned that Linux will have a lot of trouble getting into the mid-sized and small shops because it doesn't interoperate well out-of-the-box, to connect a Linux box to an AD is a total pain in the arse, serving OpenLDAP is even more of a pain.
Because Linux will run on our existing server metal, and OS X will require us to purchase new servers.
The space is automatically placed by slashcode to prevent robotic shenanigans, I believe.
If you look back, open video drivers have ALWAYS lagged months or years behind the hardware releases, it takes a long time for the developers to figure out how to access the new functions of the cards. I've been using XFree86 since the 3.2 days and I always knew that you could never run it on cutting-edge hardware. Hell, it used to be worse than not having 3D, a lot of new cards back then PLAIN WOULDN'T WORK.
As for what I linked to, those are the open drivers included with xorg, they are the drivers produced by and for the x.org community. ATI's drivers are something else altogether, and support 3D in the newer cards.
You still can do 3D on the "2D only" cards, it'll just be dirt slow because it's emulated. You'll get better 3D performance from an ancient Rage 128Pro than an X800, but the 9200 will give you the best open-source 3D experience right now.
A Randolph, Massachusetts resident going by the handle 'MarcQuadra' has discovered a briefcase that magically lowers his heating bill by 90%. When asked about it he replied only that he can now afford an LCD screen, but his eye fluids have all boiled off.
the streets in the Middle East would flow with blood to an unprecedented degree.
This is another popular misconception. Israel has one of the most powerful millitaries in the world, and they are a nuclear power; an invading army would be destroyed quite handily by IDF, and Israel wouldn't hesitate to use a small nuke on the invader's capitol. For this reason, Israel is safe from invasion. Individuals might be suicidal, but nations are not.
As for the Palestinians already inside 'greater Israel', the reason they fight with suicide bombs and crude mortars is because they have nothing else to fight with, they pose little to no threat to Israel on the whole. And also, our aid doesn't change their situation, it only aligns us with their oppressors.
We sow discontent across the world for standing unquestioningly with a fundamentalist government that feels no remorse bulldozing an entire block of houses because one has a tunnel used for smuggling arms in it. Tell me that we would accept that sort of behavior from anyone else and still call them an ally.
Our entire relationship with Israel was born out of cold war paranoia that the middle-east would go communist and a powerful domestic Jewish lobby. The population of America stands for it to this day, when the relationship serves us no purpose but to make us the target of terror, because of an irrational and media-induced American guilt for the holocaust.
I've had this argument with several of my Jewish friends, several of whom are stalwart neo-zionists. There would be no terror to wage war on if we didn't so blatantly ask for it.
As for other countries in the middle east, they're all U.N signatories, and just because we have a really big fscking millitary doesn't obligate us to make sure the maps don't have to be redrawn. Our influence on the oil producing middle east countries is arguably the leading reason those people haven't stood up and wrought governments that suit them better, our foreign policy to-date has been to line the pocket of whatever dictatioial regime will sell oil to us.
As a superpower, we should be aware that hubris will most likely be our downfall. Our foreign policy should be as hands-of as possible, let the U.N. come to decisions involving territory and soverignty, thats why it was created. If we follow our current path, we'll come head-to-head with China eventually, when they realize that we've got interests surrounding them and they stop loaning us money.
But Bin Laden doesn't WANT unorganized chaos and death. This is a common misconception of mose Americans. If you actually READ the stuff Bin Laden says, his goal is to get us and our influence out of Muslim lands.
If Bin Laden wanted to kill as many Americans as possible, there'd be people getting shot at malls and suicide bombs in America EVERY DAY. Trust me, there's a LOT of available suicide manpower here in the U.S., they just aren't tapped beause the goal of terror is to make a point and get your needs met.
We could stop AlQaeda in ONE DAY if we stopped giving Israel (a leading EXPORTER of arms) aid and a blind eye, and brought our 'stabilization troops' that prop-up the House of Saud back home. Instead we march right into the foray at great human, moral, and financial cost.
If AlQaeda made a computer virus, it would have a payload that showed messages on the screen of your machine like "Stop supporting Israel and I'll stop planning attacks." or "There were NO beheadings before Abu-Gharib, and we behead only those directly involved with the occupation" or even "Click -HERE- to see what U.S./Israeli millitary action inspired me to take revenge on YOUR towers."
I can EMULATE x86 on a PowerPC 750 and get better performance/watt than VIA's EDEN chips. Really, VIA chips have always been seriously weak sauce, just because they can't clock them well and they use not-much juice doesn't make them superior in any way.
The Pentium-M and AMD GEODE NX CPUs produce MUCH more horsepower per watt than VIA's chips. VIA's only advantage is that they mass-produce Mini-ITX boards and sell them to distributors. If someone mass-produced a GEODE NX bMini-ITX board, it would wipe the floor with the EPIA, fanlessly.
I just bought video cards a few weeks ago. the RADEON 9200 series is fully-accelerated with open-source drivers, and it doesn't need a fan, which is awesome because my system is whisper-quiet now.
It's also plenty fast for what I do with it, I'm only 2D for 99% of the time I'm on the machine anyway, and when I do play 3D games they're a year or two old, so they run fine.
Go take a look.
Cards marked with 2D only will do 3D in software (read: exceptionally poorly)
OK. YOUR card is NOT supported for 3D. You're seeing software rendering on your 9500.
n .4 .html
Please read the driver info page for the radeon at:
http://xorg.freedesktop.org/X11R6.8.2/doc/radeo
On my 7500, BZflag plays quite well with the eye candy turned on, though I prefer GLQuake.
I'm glad we're finding truth here, though I'd like to point out one more thing, I have a 'last laugh' personality. :-)
My point was that it's simply easier and more efficient to pass those offsets around in single registers and do calculations in single operations.
True, but how much LESS efficient is the other overhead of the OS when running full 64-bit? If large disk access incurs 5% less overhead, but the OS uses 20% more memory to run itself and apps, are we really gaining anything? (numbers shamelessly pulled from my arse)
I think the innovation (bus speed boosts, hypertransport, PCI-express, etc) will be happening on the 64-bit platforms now, so the reason to 'go 64' will be the better features and forward compatability, not the bitness of the system itself.
I wonder when Linux users will stop buying hardware that doesn't have published interfaces.
I also wonder when people with ATI card that are actually supported will realize it. My RADEON 9200 and 7500 get full 3D acceleration without the closed drivers.
Having lots of ram in a fileserver can help with caching/buffering of data from disk.
yes, but returns diminish after you exceed your 80/20 'working set' which at any given moment is usually under 2GB.
files of 2G or more are not that uncommon. DVD's and raw audio/video can easily produce multi-gigabyte files. For 32-bit cpu's to work on large files, various kludges and work-arounds have to be used.
Not really, 32-bit CPUs easily handle file I/O operations in the multi-gigabyte range, it's 64-bit precision calculations that 64-bit CPUs benefit from. Even if there is some overhead from dealing with huge files, the VAST majority of a file server's performance issues stem from disk and network subsystem performance. My file server regularly handles files over 4GB (DV files) and the CPU level rarely exceeds 10%, and that's on a 350MHz machine.
Before then, such files could only be made and accessed on 64-bit platforms.
IIRC, the file handling for even the 64-bit flavors of Linux were 32-bit before the mid 2.4 series. Both platforms got this at the same time.
And hard disks (let alone RAID arrays[sic]) are getting so large that the sectors can't be addressed with 32-bit integers (4G * 512 = 2T).
Sorry, that's not making any sense to me. ATA subsystems have been using 48-bit LBA addresses for a while now, and SCSI uses similarly huge numbers. You assume that each byte of the drive needs to be addressed, but this is not true, each SECTOR needs to be addressed, so you don't have to multiply by 512B/sector. Drives speak only in sectors to their controllers, it's the software component of the filesystem driver that decides what the contents of the sector will be.
Like I said, 64-bit file servers don't really have any advantage over 32-bit ones today, unless you need more than 2GB RAM. They WILL become popular before their time though, because of misconceptions like yours and the falling price of 64-bit computing in the marketplace.
I'd also like to note that I'm not against 64-bit computing, I'd rather move up before we hit the limits of today's hardware. But I think we have to be educated about what we can alreay do with 32-bit computers. I estimate that I'll be peachy-keen with my own 32-bit machines for about three more years. I'm suggesting 64-bit file servers at work now, actually, because I know that AMD's Opterons are overall better, more efficient, and cheaper than Intel's server chips; not to mention that it'll be the best way to finally get rid of our Dell servers (pray for me).
Because the mini doesn't need ANY fan at all. The mini cools by convection it it's tiny plastic shell, it'll be even cooler in a PC case.
If you place stuff correctly in a PC case you can even get the heated air from the CPU to draw in cool air over the hard drive.
Because the mini runs OS X and Linux, the Opteron runs Windows and Linux. OS X is an awesome no-hassle server platform.
Because 64-bit doesn't mean ANYTHING when you're a file server with under 4GB of RAM. I'd rather NOT have a 64-bit computer right now, I'm quite happy with 32 until there's a compelling reason to move up. File servers have no reason whatsoever to go 64-bit unless 64-bit parts becomes cheaper through volume than 32-bit.
I'm considering doing this because my 300MHz G3 tower is getting a bit long in the tooth, and I'm doing more and more with it these days. Stepping up to a 1.25GHz G4 would be quite nice, but I won't do it unless I can put two hard drives on at once.
Ahh. You CAN boot off a software RAID in OS X, but it has to be on a supported bus. The USB and FireWire busses do not have RAID drivers in firmware, IIRC. Anyone who wants to do RAID on firewire or USB is foolish anyway, the bottleneck is the bus and the 'glue' chip, not the disk.