Slashdot Mirror
If Mac OS X Came to x86, Would You Switch?
A not-so anonymous Anonymous Coward would like to put this query before you: "I'm not a fan of Windows, and never have been, but I am a fan of the x86 architecture. I really like Linux, but there are still a few issues that are keeping me from switching completely. I really like Mac OS X but I don't want to drop $2000 on a computer that is only as fast as an x86 computer at half the price. Darwin, Mac OS X's unix-ish core, has been ported to x86 and Microsoft's upcoming Longhorn OS seems to be disliked by everyone but Microsoft. If Apple released Mac OS X to compete with Longhorn, would you switch?"
Is one of "will my hardware be compatible with OS X?" -- if I could be assured that my hardware will work as well under OS X on x86 as it does under Windows XP, then I would switch in a heartbeat, or at least dual boot. Application support is another issue, as is migrating data.
This question does not have a simple answer like "yes" or "no" or "maybe" -- there are a lot of dependencies on each answer.
Never going to happen though, since Apple make their money from hardware, not the OS.
I would most certainly purchase and install it.
Doesn't mean I wouldn't still run windows. Possibly do a dual boot or a windows on mac kind of solution.
Ain't never gonna happen though. Apple makes money off there hardware and the OS is why people purchase the hardware. Be a foolish thing for Apple to do.
I can't count the number of times that I've heard this asked. The obvious answer is that yes a lot of people would switch if OS X was ported to x86. But I also can't count the number of times the people who keep asking this question have been told how irrelevant it is to do so. OS X is not going to be ported, for the simple reason that if it were Apple would go under and then OS X would no longer exist.
If you need to know why that is, just google for "if os x were ported" and you'll find the same explanation on thousands of pages. I don't feel like rehashing it here.
You're out of date. Macs are comparable (the WSJ's Walt Mossberg even claims cheaper) in price/performance to x86 boxes. When you factor in the reduction in neck pain, the lack of truly low-end macs is easily compensated for. OTOH, you can always get a used mac; OSX runs fine on any PPC version. As to your question, one of the main reasons that OSX is able to be so stable and still provide all of the eye-candy is because of a very small HCL. That advantage would be lost by moving to the rather chaotic wintel platform.
The only "issues" I have with Linux is being forced to use certain windows apps (work mostly and no, I'm not leaving a job I like just so I can delete my windows partition). I also have a Mac and yes, if MacOSX was available, I'd install it, but I wouldn't "switch" - why should I use just one OS?
The whole "switch" thing is for basic users I guess. The rest of us aren't afraid to partition a hard drive.
Furthermore, I regret to say that I don't see much prospect of any of the Linux GUI efforts approaching the ease of use and elegance of OS X any time soon - partially due to a lack of imagination, and partially due to being over-wedded to X which is evolving way too slowly and is over-wedded to a basic design that is simply outdated.
The only apps you could use would be source-based unix stuff, which you can use on linux anyway, and many of which won't actually run on OS X without a lot of work first.
So no...I wouldn't switch.
For the past several years I have been buying nothing but Apple products for myself, coworkers and family members. I have been willing to pay a higher price for better (more innovative) physical design, less cabling, and an innovative OS. The downside has been limited hardware choices that are generally a few steps behind the curve of x86 machines. That being said, I would seriously consider 'switching' to OS X on a x86 machine if it was produced and supported by Apple and possibly the hardware vendors. At the same time, however, I really like the ever evolving designs that come from Apple. When was that last compelling design change of the ubiquitous x86 desktop or laptop (maybe from Alienware)? So for me to do a pseudo-switch, I would probably also need to see some changes from the hardware vendors (how many cables run across the desk of a typical workstation?). To sum it all up, I think it would be a great thing for Apple to release OS X to the masses. It would certainly send Balmer on another sweaty spin!
My
I hate to tell you this, but the problem with x86 has always been Windows. The BSOD was not built in at chip level. I'm writing this on a Linux box that hasn't been rebooted since March.
It never ceases to amaze me how so many people who use their computers for just basic, simple tasks like office functions act like they have this incredible need for powerful hardware. I bought a Compaq Presario with a Sempron 2800, 256MB DDR Ram and a 80GB hdd for only $445 including S&H, and with SuSE 9.1 it does everything I need. It's not a game machine, so uhhh why exactly when it's just going to run Java and C++ programs for class, would I need the latest Athlon64 or Pentium4?
The reason for owning a Mac has never been about power, but utility. Every convert to MacOS X from Windows that I know switched because Macs are actually much more useful in many areas than Windows PCs. The hardest pill for many of them to swallow is that the "Apple way" really is significantly easier and more productive than the "Microsoft way."
The average computer user who could afford one, would actually be much better off with an iMac or iBook than a typical off the shelf PC. It gets the job done, and done well and it is made much better than the usual PC.
Click here or a puppy gets stomped!
drop a few hundred dollars (or pounds) on an eMac. If you find out that you don't like it flog it on ebay - mac's have great resale values. If you find that you do like it flog it on ebay and buy a more powerful model... duh!
there's no chance that apple will release OS X for X86.... and the software developers will not stand for another platform, cpu, os change.
also, the apple mac hardware would get left behind which is where apple make the money. unless of course osx86 was a poor cousin that lacked features or support and why bother in that case.
Why reinvent the wheel?
t ml
s -Are-Slo wer-Right-36964.html
Macs are More Expensive, Right?
http://www.linuxinsider.com/story/36120.h
But Macs are Slower, Right?
http://www.linuxinsider.com/story/But-Mac
Carlos
If Apple switched to x86 (well, x86-64 now), they wouldn't let it run on your commodity boxes. You'd have an expensive (although less so than PPC) x86-64 box with OpenFirmware BIOS and a few Apple ASICs to provide the same functionality that Apple has on the PPC. You'd be buying Apple hardware to run it on, without a doubt.
There are several reasons for this:
1. Apple makes a lot of money on their hardware.
2. OS X has limited driver support, opening up to all breeds of hardware would slow the development of the OS down and reduce stability.
3. There's stuff you have on Macs that just doesn't exist with your typical PC BIOS, stuff like target-mode and netboot (much better implmentation than PXE).
4. Apple are about the total experience of the platform, putting OS X on your Dell with it's rat's-nest of cabling is something that makes Steve Jobs cry. Steve has a VISION, and a huge part of it is massive reduction in cabling.
Don't hold your breath for OS X on commodity x86 boxes, it'll NEVER happen. Apple might switch to x86-64 someday if the PPC architecture hits a dead-end, but I find it more likely that the opposite is true.
I will also venture to say that the submitter of this story has something wrong with him if he prefers x86 over PowerPC. The PPC architecture is beautiful, simple, and clean. And Apple isn't the only company selling PowerPC hardware.
"Sometimes, I think Trent just needs a cup of hot chocolate and a blankie." -Tori Amos on Nine Inch Nails
You mean like the nVidia and ATI graphics chips they use?
Yes, they do control what goes INTO their machines, but it's not like they make it all, and it's not like the hardware they use is so completely unlike what's common on the x86 platform.
I think they could achive a lot by simply not worrying about supporting the kind of legacy hardware Windows always strives to, and focusing on just the common x86 hardware. nVidia & ATI pretty much cover most of the graphics chips out there on modern hardware. USB device support's already there. There's not THAT much to have to support to just work on the average desktop, and support for more exotic hardware can come later. If some hardware's too shoddy or unreliable to maintain the Mac standard, then don't support it!
There are reasons for the decisions that Apple makes, and a lot of them are documented. You can look at their Human Interface Guidelines here.
To quickly answer why they have the menus at the top of the screen: Fitts' law, which describes the amount of time that it takes to move a mouse pointer to an object on the screen, indicates that it's easier to get to the menu items at the top, since you can move your mouse pointer up with no regard for missing the menu bar. Because you run up against the edge of the screen, that parameter is effectively infinite, and reduces the amount of fine controlling you have to do to select your item.
You can also do further reading about such user interface decisions in the book 'The Humane Interface' by Jef Raskin, one of the orginal developers of the Macintosh and its interface. Their decisions for dragging disks to the trash, and having a one button mouse really DID make sense at the time, but a lot of those decisions are lost on us now, especially if we don't have a long Mac background.
Who *REALLY* is a fan of x86? BIOS should have died a long time ago. x86 is a hideous architecture with an ISA that makes all assembler code look like vomit etc. I think the only thing people like about x86 is the cost. I think the more important question is can we ever replace x86 with another architecture (I like SPARC and with Intel's R&D I'm sure we could get over any problems) but keep it as open and mix-and-match as x86 is now? Write a decent emulator and legacy code is taken care of. I'm sure OSS and it's upward trend would minimize the transition costs. But really when can we get rid of this turd?!?
Your CPU is not doing anything else, at least do something.
Just take a look here:
clicky click workgroup management
Ye needs but ask sir...
A fool throws a stone into a well and a thousand sages can not remove it.
If you look at any of the remaining NEXTSTEP/OPENSTEP archives, or search for an old OmniWeb beta, you'll find files with names like this: That NIHS.b stands for Next,Intel,HP,Sparc-BINARY.
One binary runs on 4 different architechtures. If they could do it then, with most systems (all architectures) running between 25 and 100MHz, they should be able to do it equally well now, with a narrower range of hardware to support (nVidia or ATI, x86/64 or PPC as opposed to 4 totally different approaches in respect to CPU, display hardware, bridge ASICs, etc)
There's the issue of AltiVec/SSE2, etc, but there were challenges 10 years ago, too..
>This would kill apple however because nobody in
>their right mind would pay $2000+ for a good mac >when they could pay for a PC at $1500 and get mac >OS on it as well
apple is not as expensive as it used to be you know...
today you can buy the (delicious) imac g5 with the 17'' tft screen for $1300 -- and you get a 64-bit machine. An athlon64 with similar specs doesn't cost much lower, and you don't get the all-in-one design.
And $3000 for a dual 2.5ghz,64 bit is a good price. Definetely at the low end of the dual market.
Washington bullets will simply be known as the "Bulle
I wouldn't switch, because I'm perfectly happy with GNOME running on a Linux kernel.
OSX is really pretty. But GNOME is pretty enough for me. I love the clean, tidy look, and the antialiased fonts. And I have chosen a desktop theme that I like.
OSX is really stable. So is GNOME.
If you want to try out GNOME, the best way is to install Ubuntu Linux on a spare computer. (It doesn't have to be a brand-new computer, but the older and funkier the computer is, the greater the chance of a problem.)
The easiest way to try out GNOME is to get a Gnoppix CD. You boot from this CD and it will run GNOME on a Linux kernel, without touching your hard disk in any way. So you don't risk your data. And by the way, this makes a great disaster recovery tool, even if you are a Windows user and you aren't ready to switch yet.
I'll bet there is someone writing a "KDE works for me" posting right now too. KDE is also a good environment, although I personally prefer GNOME. To try out KDE, you could get a Knoppix CD. This works the same way as Gnoppix (and in fact Gnoppix was derived from Knoppix, not the other way around).
In short, anyone who has already switched to a *NIX desktop (GNOME, KDE, Xfce, whatever) is unlikely to be tempted by an x86 OSX.
steveha
lf(1): it's like ls(1) but sorts filenames by extension, tersely
Of course, if it was free, or could be easily pirated, and it worked as well on my computer as Linux currently does, I would switch. But thats not accurate.
First of all, if it ever came to be, it would have to be hideously expensive. Don't think about the $100 cost of OS X upgrades now. Those are for people who have already paid their dues to Apple by buying a Mac. If it was any good, an x86 Mac OS port would wipe out a fair share of Apple's current hardware sales. Even if they could, say, double their current OS market share by running on cheaper commodity hardware, they would still need to make half the profit on each new, non-upgrade copy of x86 Mac OS that they currently make on the average new Macintosh sold. I would be very surprised if they could make this up with a retail price under $400. I definitely wouldn't pay $400 or more for it, as slick as it is, and compared to free Linux and "free" preloaded Windows I doubt many other current PC users would, and it would never be a market success for Apple if only existing Mac users bought it.
And thats all supposing that the product is every bit as good as the current version of Mac OS for Apple hardware. That means that they would have to support seamlessly every possible combination of PC components that could show up on a computer made in the last 3 or 4 years. Darwin x86 certainly can't do that now, and even if they could port over every current FreeBSD driver, plus support every video chipset they support on Macintoshes now, it would be far from universal, although it would be good enough for me. Microsoft spends a lot of money on testing and driver development to ensure Windows works on every wacky system they claim it will run on, and that is even given that most of the device drivers are written by vendors.
Given these constraints, I don't think Apple can bring a viable x86 Mac OS port to market at a price low enough to be successful, so no matter how cool you think it would be to have OS X on your computer, it isn't going to happen.
"(Man) tries to live his own life as if he were telling a story. But you have to choose: live or tell." --Sartre
Yeah, you're right, with the exception of a few CPU-level bugs along the way, the BSOD hasn't been built into the CPU, but that's not to say that it's always Windows' fault.
Other things that go into the Windows world's instability include:
El-Cheapo hardware du jour. This includes many, many x86 mobo manufacturers, as well as bottom of the barrel RAM and PSU suppliers. Guess what: If you're truly talking about making something the same as a $2,000 Mac for half the price (hyperbole, I know), then you're engaging in some of this, and it is where a lot of the BSODs originate.
As a follow-up to section 1: shitty driver support, particularly in the 9x days when everything, not just video, had an easy chance to cause system-level problems.
When people say x86 in a debate such as this, they generally mean the platform as a whole, not the cost of the chip. A Pentium 4 chip by itself is as useless to me as a G5 by itself. But to say that Linux or some other non-Windows OS is going to be magically immune to the cheap-ass, no-QA hardware that you frequently encounter in the x86 world is completely off base.
Tiny cache? x86 processors have larger L1 caches than most other processors. The L2 cache size is usually pretty puny though, and few PCs support L3 cache (the AMD K6/2+ and K6/3+ had onboard L2 and the system L2 became L3, that's all I can think of in that category just now.) AMD made several very nice x86-compatible processors. The lack of registers is pretty lame but it is addressed in x86-64 and for many programs it is not that big a deal anyway because both intel and AMD (and their assorted competitors, as well) have implemented register renaming and other tricks to improve the speed of context switches and the shuffling of data due to register starvation.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
As a techie, I always get bombarded with "Where can I get a good computer?" questions, and I just tell people either at Best Buy for 500 bucks or a Mac.
A few weeks ago someone answered that they don't like Macs because "they like to have control that Windows gives them".
So, yesterday that person needs help burning pictures off the digital camera to 3 CDs. It took us fourty (40!) minutes to burn 600 megs of data on a well equipped Windows XP machine (3.06GHz P4 HT, 512 MB RAM). This is why:
I drag the first set of files onto the CD, they burn OK (albeit a bit slow).
I drag the second set of files and get an "incorrect function" error. I'm thinking WTF?
I use a new CD and some drag-to-disc program comes up and burns the pictures much faster than the first CD did.
I try to burn the final CD, but get the incorrect function error again. It took me 20 minutes of CD swapping, ctrl-alt-deleteing, and cussing to figure out that I had to right click on the burner icon and enable CD burning for it.
Well, duh, one might say, of course you have to have the CD burning enabled. You might think so, but you'd better not wonder why the first two CDs burned, but the last one required enabling.
What I'm trying to say is that in the time it took me to burn 600 megs of data on a very well stocked Windows XP PC, I could have had the very same pictures sorted into albums, posted on my website (which runs on the same machine), and burned on a cd on a 1.4GHz G4.
Apple way might be more expensive at first, but it doesn't require you to randomly click on things to make them work. (Provided that the PC has a slew of 3rd party applications to keep it working in the first place.)
If you don't know what AltaVista is (was), get off my lawn.
It isn't the integration that causes requirements for reboots.
It's the base concept of not being able to move or rename a file once it is open. This is the same whether you are talking about a simple text file, or your system files for the OS.
Most filesystems (fat, ntfs, and of course all unix filesystems) have an abstraction between a directory entry, and a filesystem wide number that identifies a file -- in unix filesystem terminology that is the "inode". The problem is that windows does not use that natural abstraction.
In unix, if a program has a file open, you can "delete" that file -- wait it's not gone yet[1]. What happens under the covers, is that your filename in a directory points to an inode. Each inode has a count of the number of references to itself. When you open the file that reference count for the inode is incremented. So if there is one program with the file open, and the inode is referenced in one directory, the reference count is now two. Once that count goes to zero, the inode is unlinked -- which means deleted, but since "delete" is multi-step in unix, you need more terms.
[1] This is a great way to work with temporary files -- once you open and delete it, nobody else can access the file and many security threats with temp files are completely avoided.
Now you have a system file that needs updating. You delete the file (which just removes one reference from the directory) and the system still holds a reference to the inode and continues operating as before. Then, you write the new updated file to the same directory with the same name, but it doesn't cause a conflict because the new file goes into a different inode. Once the files are replaced, restart the individual application and the update is finished.
On windows, to update a system file that can not be closed during the operation of the system, you put the files in a special location with a script that specifies their desired location. Every time windows boots it runs those scripts that replaces files before they are opened. That is why windows will always require more reboots than unix based OSes.
There is more to updating system files than that of couse. For instance, most unix servers do not run a graphical environment on the server and every version of windows since NT boots into a graphical graphical environment. Most Unix based systems use the Xwindows system for their graphical environment, and an update to that environment only requires a restart of Xwindows not the entire system -- which is important if you are running any services on your computer that people depend upon.
There: Something at a specific location.
Their: Owned by someone.
Please make sure your english compiles.
I have found in corporate environment where the user is regulated from doing anything stupid, the most common failure is hardware. One of the sales computers blue screen this morning. Do I blame Microsoft for the BSOD, even though it would give me brownie points here, after reviewing the machine in the lab, I put the blame on the RAM maker, simply replaced the stick and the employee was back working.
Just because a windows box fails doesn't mean it's Microsoft's fault.
Now to answer the article's question, I wouldn't switch to OS X, but I would probably buy a copy and run it in the lab. If I find uses for it, I may deploy it, but in the end, an operating system is a tool to get a job done, you simply use what you deem to be the best tool for the job. Which for me, in most cases is Windows, though I use *nix variants for certain tasks, where Windows is just not cost effective.
The entire GUI and all of the hundreds of libraries ("Frameworks" in Mac OS X) that Mac OS X apps depend on would need to be ported, and many of these are only designed to work on PowerPC currently.
Well, no, that's not true. The actual OS X GUI, frameworks, and libraries are largely NeXTStep, and that stuff is quite portable and even ran on x86 at some point.
Of course, OS X also has Carbon and the backwards compatibility stuff in it and that might be harder to port.
Apple cannot do this. For multiple reasons that have nothing to do with technology and everything to do with marketshare and market forces.
.Net stuff, but here again, who would actually buy this? .Net is far enough advanced that it is the king of Windows and no big shops would move to Cocoa.
If Apple were to switch, it would start at point zero, with no applications, just as it did in 2000 with the Mac OSX public beta. It would not even have the Classic layer to run the old Mac apps, although that is less of a concern these days. These days Mac OSX on x86 would either have to have some very fancy easy to use WINE like environment, so that Windows users could use their current apps until OSX equivalents came along, or run Windows in a VMWare like environment.
Also, you would have to kiss MSOffice for OSX goodbye. You, as an OOo user might think this is insignificant, but it sure as hell isn't to those who use it.
Then, there would be the hardware problem that people like the article author always always forget. Apple makes most of its money from hardware. Do you think Apple would still make as much money and sell as many units if the OS was able to run on x86 commmodity hardware? Obviously not. That would force the price of the OS up. How many newbies would then pay for the OS?
Talking of newbies, how many of the gazillions of windows users who currently have never heard of OSX, or think OSX is still the same crap they used back in 1995, will fork over $100 plus for an OS without Office, without games and without other pro applications? I seriously doubt that Apple's pro apps like FCP and Logic alone are enough to sustain the platform, not to mention the intensely pissed off current users on PPC.
That means that the only people buying OSX on x86 would be geeky types like those here who are so fucking stingy (and I don't mean poor - I'm poor yet I use a Mac because it's so good) that they bitch about a $100 price difference in a computer. The vast majority would not use OSX or even think about switching.
Apple's only recourse in this case would be to make an x86 mainboard, using a special OpenFirmware with no bios, such as current Mac mainboards do, to make it incompatible with other x86 machines so that it would not encroach on Microsoft's Windows territory too much and so that it would keep users from using other x86 hardware.
And the advantage of that over its PPC platform is right around zero, so why even think about it.
Apple could have done this exactly once in the past. Back in 2000 when everyone was started to switch to OSX and there were no OSX applications for the new OS, Apple could have gone with a proprietry x86 motherboard and kept on producing a few last PPC machines until their classic MacOS users had switched and there were enough new OSX apps. That time passed as soon as new apps started coming out for OSX and people started investing money in them.
The only thing Apple could possibly do to make money on x86 these days is possibly port its Cocoa frameworks and devtools over to windows to compete against MS'
I am so so fucking tired of some one night wonders asking this same stupid question, when it is so obvious that it is just a geek cheapskate wet dream.
If you can't utilize the full-power in a machine, or if you will not utilize said power in your everyday operations, then all power in excess of what will be utilized is useless power. Now, on a rather different topic, if you're trying to suggest that a $1300 Athlon system (assuming that you're going for the best cost/parts ratio you can find and consequently build it yourself) is comparable to an equivalently priced imac G5, then I'm going to have to strongly disagree. Now, I am not including the price of a monitor, which, assuming you don't already have one, would be an additional cost in the Athlon system. However, in terms of both upgradeability and utilizable performance in currently available consumer-oriented games and applications, I would suggest that the Athlon system as stated above would absolutely kick the teeth out of the imac. If you want to talk about the advantages of the Mac OS, talk about the OS, but don't for a minute suggest the price of hardware for Macs is somehow equivalent to PCs. The upgradeability issue alone shows that PCs win hands-down in this category. An example is in order (fast-forward to this hypothetical future scenario): If I want to do a major hardware upgrade on an imac I purchased a year ago, I'm basically out of luck. Thus, the most cost and time-effective solution is to essentially get a new machine. With a PC that I assembled a year ago, I can easily take whatever component out that I want and replace it with an updated component. This long-term cost factor must also be considered when comparing the differences between the two platforms. Thus, in my estimation, the PC is still the clear winner as far as hardware is concerned in the lower-cost categories.
And all our yesterdays have lighted fools The way to dusty death. --Will
Most Mac users are completely disconnected from the larger PC market. Apple's prices have been the same since the mid-90s, so they assume that's how the rest of the world works.
This is completely wrong. I am looking for something to do digital video on and the new G5 iMac is very competitive. With 1GB Ram, 250GB HDD, DVD burner, 20" LCD and Final Cut it is $2300 and this includes a high end graphics card, TV out and some nice software. For comparison a Dell Precision with 1GB Ram, 250GB HDD, DVD burner, 20" LCD, Adobe Video Collection and graphics card is $3940
Of course the Dell Precision is extensible which makes it more attractive but if you can get a machine with the same spec as the G5 iMac for less, good luck. Its easy to say that Macs are crazy expensive and Apple is still in the '90s but the facts don't bare this out at all.
- Macs don't have a $2000 start up price tag, they actually start at $799 with the eMac, for portables $1099. Those of which are better spec'ed than low-end PCs.
- The second point is that there are no more G3's from apple, it's been that way for a while, plus no apple computer is spec'd below 1GHz.
- You can not get any apple computer without firewire. (It's an odd one I know, but I sometimes see some rattle about firewire cards and macs.)
- OS X, runs smoothly on a G3 700MHz, it runs smoothly on a G4 400MHz, a G4 1GHz won't leave you waiting in any application including Alias' Maya. Hence you don't need a dual 2.5GHz G5 to 'test' OS X, a second hand mac is usually just fine to try it out. (If you look hard enough, there are people giving away old powermac G4s.)
The final mistake, which is a good one, is that developers have not ignored the 64 bit G5, the reality is that developers have embraced it, and in cross platform apps, it's actually been the PC waiting for 64 bit updates from vendors such as Adobe, this when adobe apps were updated long ago for the G5.
The logic that Apple computers wouldn't sell, if PC's ran their software is also another fallacy. PC's already run iTunes and iPods, this hasn't affected the apple market share (it's actually grown because people aren't taboo to the brand anymore).
Additionally apple don't have any monopoly on the OS market or the hardware market, anyone that doesn't -want- to buy an apple, simply doesn't, it's not like windows software on x86. In a world where large purchases are governed entirely from the bottom line, you'll find that OS X on Intel would have little effect on apple computer sales.
Likewise my OSX has _never_ crashed, but I did get to visit all those bad sites, treaded all over the shop like a drunk, and didn't have to bother with a NAT, firewall, virus scanner, heck i randomly open attachments full of windows virii, I'm raw on the Internet and nothings gotten in.
Meanwhile, the PPC almost always has 64K L1 cache (32K instruction, 32K data).
The PPC 750 (the "G3") had 512k or 1MB of L2 cache running at bus speed (66 or 100 MHz). (Source: Apple - Note: This page shows L2 cache to be 512k and 300MHz to be an available option. The 300MHz PPC750 had a 1MB L2 cache).
Early G4's had 1MB of L2 running at bus speed (100 or 133 MHz). (Source: Apple)
Later G4's had 256K of L2 (at processor speed) and 1MB or 2MB of L3 (at 1/4th processor speed). (Source: Apple - 1MB - 2MB)
The G5's have 512k of L2 running at processor speed. (Source: Apple)
And lots and lots of registers...
I can understand why you might like cheap/fast-enough/common x86 processors. I just don't understand why anyone would defend the poor design decisions that bite them in the ass.
Nevermind that you're comparing an iMac to a corporate-class workstation. If you'd configured a Dimension instead of a Precision, you'd probably not have had anything to post about, sorry to say. I just got done refuting someone else's claims that "Apples aren't any more expensive than similarly equipped PCs" on The Inquirer. Oddly enough, they did the same thing you did: went straight to Dell, picked a corporate class machine. Sadly, the other people made some gross misjudgements to jack up the PC price even more, by configuring the Dell with twice the RAM, more addons, etc... But still, when you get down to it, you're comparing Apples and oranges. Back on topic, even if Apple did release OSX on x86 architecture, it'd likely run poorly (at least as compared to OSX on a Mac). It's just not in their best interests to release a "perfect" version of OSX on x86 architecture, simply because I doubt Apple makes its money selling software. If there's no compelling reason for people to buy the machines (style only goes so far, OS and other software tow the load), there's very little profit potential for Apple in releasing OSX on PC hardware.
$120 DVD-+RW drive (one of the new dual layer capable ones)
$180 Asus Mobo with built-in Optical audio, gigabit, a normal ethernet connection, AGP 8x PRO, pci express, firewire, and USB 2.0 and an Athlon XP 3200 installed
$120 120 GB harddrive
$250 Decent monitor - 19" or higher
$100 512 MB RAM
$20 Modem
$80 video card
I'm sorry.
Rant mode on.
I too used to think lists like this made some sort of sense. But they don't. The only way this list makes sense is if you are a computer expert who uses Linux (or pirates Windows), AND you are building this computer for yourself, AND you can get the parts all from the same place or inexpensively shipped, AND you somehow managed to get all the right parts out of the dozens of different types of cases, motherboards, and CPUs, AND you know how to put them together correctly AND all the miscellaneous cheap parts you bought all work together and none are faulty and need to be replaced AND etcetera ad infinitum...
Everyone who makes a list like this is just wasting his time. The list only makes sense to you, unless you are in the habit of spending many hours ordering parts and building computers for other people FOR FREE... If you wanted to make any profit at all by selling that particular system to someone, there's this little thing called economics that gets in the way. You'll have to double the cost of all the parts, add compensation for shipping, and then bill for labor. Suddenly you have a machine that costs double the price of a low-end Mac (eMac) and you still haven't even added any software!
And lets be realistic here, 80% of the populace is either not ready for Linux or simply won't be happy with it (in its current state). Even if you do discount the cost of the software, to duplicate the usability of a Mac or even Windows you or someone else will have to spend X amount of hours setting this system up so all the hardware works and all the necessary software is installed to do the same basic tasks that can be easily done on either Mac OS X or Windows.
Just as a little experiment, why don't you go ahead and build yourself one of these systems from scratch. Just be sure and keep careful track of all the time you use from planning to ordering to assembly to installation and configuration of the software. When you're finally done and have a moderately usable system, multiply the number of hours it took you by whatever you get paid at your current job. Even if you don't get paid much you will probably be shocked at the number that comes up. Your time just spent finding each part will probably negate the "$40 less" that you think you can get all those parts for.
If you are a geek who uses Linux and you have free time on your hands then building a computer from scratch can be an entertaining, educational and satisfying experience. It's loads of fun, I've even done it myself a few times. But trying to compare this list to the actual market price of any pre-built computer with all necessary software installed... it's just insane. Another thing, you do not need a $2000 Mac to get the same functionality, usability and apparent speed as a PC half the price, this is a myth today. Macs simply cannot be directly compared to PCs speed-wise for anything but a few specific operations that only a small segment of the population uses their computer for. For most poeple, a Mac that costs $800-$1200 will be more usable, safer to use on the Internet, and more fun to use than any PC in the same price range, no matter if it's running Windows or Linux.
End of rant. I hate older Macs and Mac OS 7/8/9 and have been running Linux at home for years, so don't even think of calling me a Mac zealot. I'm just tired of seeing these BS "cost" lists being taken seriously here. The general reader and the moderaters need to come to the realization that they literally make no sense for anyone besides the original poster, unless the value of your time is zero and you already have all the requisite knowledge.