G4 processors in an SMP configuration can talk to each other on a separate (and of course much faster) bus than the system bus. Someone just needs to make a logic board that supports it now.
Take a look at the specs for the Sawtooth logic board in the two high-end models. 128 bit path to memory. This thing has a system bus nearly as fast as an SGI Octane! This is some serious shit. Apple also claims that it has PCI access twice as fast as the B&W G3. The B&W has 64 bit 33MHz slots.
This thing is more targeted at SGI's visual workstations than the typical PC. That's where the price comes from. Well, all except the "low-end" model. It uses the same logic board as the G3 did.
Only problem is, Apple is moving too fast. Good support for all the stuff in the G3 still isn't done, and the G4 is already here.
There's more to screens than resolution and size. From what I've heard this screen is amazingly bright, and has amazing contrast and color. I've heard from a graphics guy who saw one in person that this is the first LCD he'd actually use for his work.
The G4 and the Cinema display aren't translucent anymore. They just have a couple of clear parts. MacDesktops has some nice high-rez pictures from different angles.
That's exactly it. The reject rate on large LCD screens in amazingly high. Every person in the graphics industry I know wants one of these. There is no way they're going to be able to make enough, even only selling it with the G4 and only from the Apple store. The only thing that bugs me about it is it means I can't go down to the local Mac place and drool over one.
I feel your pain. I just spent $2K on a G3/400 a month ago. I expected these to be coming out later, and I was going to wait, but my trusty old Mac clone's logic board died, so I figured I'd go ahead and get a G3. Well, processor upgrades should be out soon .
It should be noted that the 400MHz G4 is just a G3 with a G4 chip stuck in it. It doesn't have the wider memory bus, AGP, Airport support, faster PCI, or digital video out that the 450 and 500 will have, so I don't feel _that_ bad.
Are you talking about Mac OS X Server, which is nothing like the client version will be (and nothing like future versions of Server will be), or Mac OS X Client DR1, which is an extremely early developer release that you are violating your NDA by talking about?
Point is, making judgments about OS X using either of these as a guide is worthless.
I don't see why they're implementing MDI in Linux. I've heard the BeOS version doesn't have it (I've never seen a BeOS app that did), and I'm almost positive the Mac OS version won't have it (Mac OS won't even let you do it, you'd have to write your own windowing system), so why are they forcing it on Linux users?
Oh well. Doesn't matter to me as I'm sure they'll never release PPC binaries! Arg! I hate that.
You need to know some tricks to use the Mac OS's method well. The most useful is that if you hold down the option key while switching apps, the one you are switching from hides itself. I regularly run 10 or more apps in Mac OS (with multiple windows open in each) like this with no problems on a 17" screen. If things in the background are hidden you don't even know they're there, but choose them from a menu and all their windows just show up and pop to the front.
MDI is one of the major reasons I don't like Windows. BeOS and Linux do better, but if you do away with MDI it makes sense to put the menubar at the top of the screen rather than in the window, which Linux and BeOS (generally) do not do, though KDE supports it. Having the menubar at the top of the screen has the added advantage of making it easier to hit; just head for the top of the screen and then position horizontally. Much faster than hitting a 16 pixel high target.
I've heard that you can unplug a Firewire drive while it's being copied to and the copy will pick up when you plug it back in (although technically you aren't supposed to unplug a drive without unmounting of course). Firewire also allows for more than one computer on the bus as well, so they can share devices (not sure if that works with all devices; I assume it depends on the controller chip and the drivers). And of course unlike USB and USB-2 (is Intel still actually going ahead with USB-2?) Firewire works with _no_ computer on the bus as well, so it can be used in consumer electronics.
Re: Cost of Firewire: In theory it'll be nearly as cheap as USB and much cheaper than SCSI. It just has to get into widespread use. That's why seeing it in new computers is so good. Hopefully Apple will also put it on the new iMac as well. Anything that's going to make Firewire peripherals more common and cheaper is fine by me.
There are significant advantages to designing the OS and the hardware for each other. It's the reason the Mac has never had many of the sorts of problems with peripherals Wintel has.
Of course there are significant disadvantages as well, which I'm sure I don't need to list.
And that's quite different from Mir. It costs ~$10K/pound to put stuff in orbit. If a single ton of Mir can be reused, that's ~$20M saved!
Of course it might cost more to keep in in orbit than it would to send up materials we need. It depends on how much of Mir could be reused and how much it would cost to to the work in orbit to salvage it. I'm sure NASA and other involved parties have done the math.
You don't need many people at all in fact. Genetic material for millions of people can be kept on ice and you can send a population of several dozen to colonize a planet. There are other ways to solve the genetic diversity problems as well. If one generation clones itself to create the next there's no problem with inbreeding. With a better understanding of genetic engineering, it can also be used to insure that the even if the genetic pool is too small, none of the problems that generally associated with that actually become manifest.
Where do they draw the line? Does this apply to hard drives? What about Zip disks? What about DAT tape? What about Sony Mini Disks? What about flash cards?
It seems to me there's no clear place to draw the line here. Just about any type of digital storage can be used to record video and audio data. Hell, you can store the stuff in DIMMs.
Mac OS X isn't going to be just another Unix* OS. It's going to be the only one that's fit for use by the general public (read: iMac users and such). Linux will be there in a couple of years, as GNOME and KDE get better, but OS X is going to be there in 6-10 months. That's a _very_ big deal. Ease of use is one of the two major reasons Linux hasn't killed Windows yet. The other of course is apps targeted at the desktop market, and Mac OS X will do better than Linux there as well in the near future. And OS X hasn't really had any features dropped from it from the original announcement. They killed the x86 version, when it ships it'll have a _larger_ feature set when it ships than what was originally announced.
As for Jobs missing this whole thing, he likely knew about it before IBM even told the public. The question is what he's going to do about it. It's easy to say that Evil Apple will do the Evil Things, but I wouldn't bet on any predictions about what Jobs is going to do.
Mac OS no longer has ROM dependencies on the latest hardware (a ROM image now sits in a file on the HD and gets loaded into RAM during boot). However, Mac OS 8.5 has major issues on CHRP hardware or doesn't work at all.
Mac OS X and Mac OS X Server are different though. Unless Apple specifically codes them to check for Apple hardware they'll run with no help from Apple, thanks to being open source at the core.
Mac OS X has not ROM dependencies and it's entire core is Open Source, meaning that unless Apple specifically takes action to prevent it it's going to run on these things with no help from Apple at all. As for the current Mac OS, well, who cares? It sucks. Mac OS X is the real future of the Mac anyway.
This is great news! It looks like this open PPC platform might actually take off. Not only does this provide a (relatively) cheap alternative to Intel hardware for Linux users, but it could be the first step to an open Mac platform. Once Darwin gets ported to these things (and it's already being talked about), there shouldn't be any reason the rest of Mac OS X won't run on them. This could allow for a return to Mac cloning without the problems it had the first time. Apple wouldn't be subsidizing hardware development this time, and Apple wouldn't have to insure perfect compatibility with other company's computers (they'd have the source to the entire core of the OS, they could do it themselves).
If Apple delivers as promised with Mac OS X, it's going to be an OS as robust (or more!) than Linux that's easier to use than Windows. If it actually runs on an open platform, it could cause a bit of trouble for certain software companies that nobody likes. And anything that causes MS to lose some of it's iron grip on the OS market can't help but be useful to Linux.
If Apple doesn't screw this up it could be a major boon to Apple, Open Source software, and consumers. Please Apple don't screw up again!
Slashdot Mirror
SMP fun fact #2: MaxBus.
G4 processors in an SMP configuration can talk to each other on a separate (and of course much faster) bus than the system bus. Someone just needs to make a logic board that supports it now.
--
Take a look at the specs for the Sawtooth logic board in the two high-end models. 128 bit path to memory. This thing has a system bus nearly as fast as an SGI Octane! This is some serious shit. Apple also claims that it has PCI access twice as fast as the B&W G3. The B&W has 64 bit 33MHz slots.
This thing is more targeted at SGI's visual workstations than the typical PC. That's where the price comes from. Well, all except the "low-end" model. It uses the same logic board as the G3 did.
Only problem is, Apple is moving too fast. Good support for all the stuff in the G3 still isn't done, and the G4 is already here.
--
There's more to screens than resolution and size. From what I've heard this screen is amazingly bright, and has amazing contrast and color. I've heard from a graphics guy who saw one in person that this is the first LCD he'd actually use for his work.
--
The G4 and the Cinema display aren't translucent anymore. They just have a couple of clear parts. MacDesktops has some nice high-rez pictures from different angles.
--
As an owner of a G3 I can state that the Rage card it ships with does not have anything but a VGA connector.
--
That's exactly it. The reject rate on large LCD screens in amazingly high. Every person in the graphics industry I know wants one of these. There is no way they're going to be able to make enough, even only selling it with the G4 and only from the Apple store. The only thing that bugs me about it is it means I can't go down to the local Mac place and drool over one.
--
You can't get an SGI with a screen as nice as that 22" LCD.
--
I feel your pain. I just spent $2K on a G3/400 a month ago. I expected these to be coming out later, and I was going to wait, but my trusty old Mac clone's logic board died, so I figured I'd go ahead and get a G3. Well, processor upgrades should be out soon .
It should be noted that the 400MHz G4 is just a G3 with a G4 chip stuck in it. It doesn't have the wider memory bus, AGP, Airport support, faster PCI, or digital video out that the 450 and 500 will have, so I don't feel _that_ bad.
--
1) AltiVec is very important to Apple
2) Mac OS X is very important to Apple
3) gcc is the compiler of choice for Mac OS X
You draw the consusion.
--
Some folks think it might just have a bit more to do with Intel's multi-million dollar investment in Be than anything Apple did.
--
Are you talking about Mac OS X Server, which is nothing like the client version will be (and nothing like future versions of Server will be), or Mac OS X Client DR1, which is an extremely early developer release that you are violating your NDA by talking about?
Point is, making judgments about OS X using either of these as a guide is worthless.
--
Uh... You do realize the apple-with-a-bite-out-of-it _is_ a religious reference... Think Tree of Knowledge and all that.
Never heard of anyone being offended by it, but I'm sure some people are. Just about anything you can thing of offends someone.
--
I don't see why they're implementing MDI in Linux. I've heard the BeOS version doesn't have it (I've never seen a BeOS app that did), and I'm almost positive the Mac OS version won't have it (Mac OS won't even let you do it, you'd have to write your own windowing system), so why are they forcing it on Linux users?
Oh well. Doesn't matter to me as I'm sure they'll never release PPC binaries! Arg! I hate that.
--
You need to know some tricks to use the Mac OS's method well. The most useful is that if you hold down the option key while switching apps, the one you are switching from hides itself. I regularly run 10 or more apps in Mac OS (with multiple windows open in each) like this with no problems on a 17" screen. If things in the background are hidden you don't even know they're there, but choose them from a menu and all their windows just show up and pop to the front.
MDI is one of the major reasons I don't like Windows. BeOS and Linux do better, but if you do away with MDI it makes sense to put the menubar at the top of the screen rather than in the window, which Linux and BeOS (generally) do not do, though KDE supports it. Having the menubar at the top of the screen has the added advantage of making it easier to hit; just head for the top of the screen and then position horizontally. Much faster than hitting a 16 pixel high target.
--
BeOS doesn't support those machines.
I've heard that you can unplug a Firewire drive while it's being copied to and the copy will pick up when you plug it back in (although technically you aren't supposed to unplug a drive without unmounting of course). Firewire also allows for more than one computer on the bus as well, so they can share devices (not sure if that works with all devices; I assume it depends on the controller chip and the drivers). And of course unlike USB and USB-2 (is Intel still actually going ahead with USB-2?) Firewire works with _no_ computer on the bus as well, so it can be used in consumer electronics.
Re: Cost of Firewire: In theory it'll be nearly as cheap as USB and much cheaper than SCSI. It just has to get into widespread use. That's why seeing it in new computers is so good. Hopefully Apple will also put it on the new iMac as well. Anything that's going to make Firewire peripherals more common and cheaper is fine by me.
There are significant advantages to designing the OS and the hardware for each other. It's the reason the Mac has never had many of the sorts of problems with peripherals Wintel has.
Of course there are significant disadvantages as well, which I'm sure I don't need to list.
3 hours for the recycled can, not 3 weeks.
And that's quite different from Mir. It costs ~$10K/pound to put stuff in orbit. If a single ton of Mir can be reused, that's ~$20M saved!
Of course it might cost more to keep in in orbit than it would to send up materials we need. It depends on how much of Mir could be reused and how much it would cost to to the work in orbit to salvage it. I'm sure NASA and other involved parties have done the math.
You don't need many people at all in fact. Genetic material for millions of people can be kept on ice and you can send a population of several dozen to colonize a planet. There are other ways to solve the genetic diversity problems as well. If one generation clones itself to create the next there's no problem with inbreeding. With a better understanding of genetic engineering, it can also be used to insure that the even if the genetic pool is too small, none of the problems that generally associated with that actually become manifest.
I should buy some land 100 ft from the border on the US side and sell recordable digital media! :-)
Where do they draw the line? Does this apply to hard drives? What about Zip disks? What about DAT tape? What about Sony Mini Disks? What about flash cards?
It seems to me there's no clear place to draw the line here. Just about any type of digital storage can be used to record video and audio data. Hell, you can store the stuff in DIMMs.
Mac OS X isn't going to be just another Unix* OS. It's going to be the only one that's fit for use by the general public (read: iMac users and such). Linux will be there in a couple of years, as GNOME and KDE get better, but OS X is going to be there in 6-10 months. That's a _very_ big deal. Ease of use is one of the two major reasons Linux hasn't killed Windows yet. The other of course is apps targeted at the desktop market, and Mac OS X will do better than Linux there as well in the near future. And OS X hasn't really had any features dropped from it from the original announcement. They killed the x86 version, when it ships it'll have a _larger_ feature set when it ships than what was originally announced.
As for Jobs missing this whole thing, he likely knew about it before IBM even told the public. The question is what he's going to do about it. It's easy to say that Evil Apple will do the Evil Things, but I wouldn't bet on any predictions about what Jobs is going to do.
*Yes, I know it's not Unix(tm).
Mac OS no longer has ROM dependencies on the latest hardware (a ROM image now sits in a file on the HD and gets loaded into RAM during boot). However, Mac OS 8.5 has major issues on CHRP hardware or doesn't work at all.
Mac OS X and Mac OS X Server are different though. Unless Apple specifically codes them to check for Apple hardware they'll run with no help from Apple, thanks to being open source at the core.
Mac OS X has not ROM dependencies and it's entire core is Open Source, meaning that unless Apple specifically takes action to prevent it it's going to run on these things with no help from Apple at all. As for the current Mac OS, well, who cares? It sucks. Mac OS X is the real future of the Mac anyway.
This is great news! It looks like this open PPC platform might actually take off. Not only does this provide a (relatively) cheap alternative to Intel hardware for Linux users, but it could be the first step to an open Mac platform. Once Darwin gets ported to these things (and it's already being talked about), there shouldn't be any reason the rest of Mac OS X won't run on them. This could allow for a return to Mac cloning without the problems it had the first time. Apple wouldn't be subsidizing hardware development this time, and Apple wouldn't have to insure perfect compatibility with other company's computers (they'd have the source to the entire core of the OS, they could do it themselves).
If Apple delivers as promised with Mac OS X, it's going to be an OS as robust (or more!) than Linux that's easier to use than Windows. If it actually runs on an open platform, it could cause a bit of trouble for certain software companies that nobody likes. And anything that causes MS to lose some of it's iron grip on the OS market can't help but be useful to Linux.
If Apple doesn't screw this up it could be a major boon to Apple, Open Source software, and consumers. Please Apple don't screw up again!