“They started queueing at 2.30am. By 8.30am the line continued for hundreds of yards, snaking down an escalator, under a road, up another escalator and along the other side of the street. The first MacWorld Expo of the year had arrived at the Moscone Center in San Francisco. And chief executive Steve Jobs’ keynote speech was the hottest ticket in town.” Read Guardian’s and Ars’ report from MacWorld.
This article is all we need for another 50 flame-posts.
at this MacWorld. I always wanted to go. Guess I’ll wait for the one is Boston(?). Mac users are a phenomenon.
Oh well, I was really hoping on an anouncment of OS X on x86 going out for public distro, but I guess Apple are holding on to every straw of hardware dependancy they can get… Sad really, considering how damning this report really is of their current business model.
HOPEFULLY they wake up soon.
You are indeed a Lone sad OSer that won’t get what he wants anytime soon..
and nothing to show for it..
Apple may be good when it comes to making useful software, their Keynote software and other i-products certainly are examples of focusing on the needs of the user.
But I dont get it, why pay for a mac? Over the years I have worked with several macs, beginning from the 8500, and all the way to the newest powerbooks (ahh, second newest).
I never felt religously connected to my mac, in fact I hated every second of it. MacOS 9 just blows up when asked to perform some real work..
Mac OS X gets in the way, too many windows all over the place.. Dock gets in the way… I tried hard to like it, but failed.. I missed BeOS and workspaces and tracker… Left mac for Linux/KDE and learned to live with several desktops…
But I never understood what made a desktop mac an interesting buy? (except for the cube, and shuttle has now caught that market nicely).
A powerbook I can somewhat understand, their titanium frame and sleek design are as of yet unmatched by wintel-laptops…
Can anybody give me one reason for a “normal” developer to buy a desktop mac? and no, putting linux on it doesnt count.
BSD Unix/X11
free and excellent developer tools, e.g., Cocoa. Also Applescript
Also, wait until the 970 machines. This is the low point for the pro desktops.
I use a 4 year Old PowerMac and I upgraded it recently for about the same price of a new PC. I’ve probably paid what most PC users would’ve paid over the past 4 years for two computers on my PowerMac G4, and it’s still running OS X beautifully.
Stop bitching about OS X for X86 buy yourself an older Mac or a newer one and I guarantee that you’ll be pleased for the next 5 years, rather than owning a PC and bitching for the next 5.
>rather than owning a PC and bitching for the next 5
I don’t bitch about my PC. It works beautifully. I bought it in 1999, and even back then was not top of the line (dual Celeron 533, 66 Mhz 256 MB SDRAM). It runs great all the 8 OSes i have there installed currently, including XP PRO.
My husband owns a dual PIII at 450 Mhz, and he doesn’t want to upgrade either. XP PRO works for him greatly as well (he also bought it in 1999, with 512 MB RAM).
On the other hand, I do own a Cube 450 Mhz with 1 MB of L2 cache and 448 MB SDRAM 100 Mhz (effectively, at least as fast as my dual Celerons, maybe more!), however OSX does not work as fast as any of my OSes on my x86 box. Not even as fast as Linux (let alone XP PRO and BeOS :-). However, indeed, the user experience with OSX is better than in any other Unix, or Linux.
1) Motherboard with proprietary apple boot rom. No boot rom, no OS X.
2) Short list of supported perfierals. To ensure the the best experience for the user, Apple will make no attempt to support every manufacturer of every widget.
3) No retroactive support for your existing components from companies. Where’s the $$$ for them to write OS X drivers for your existing PC? Uh-Uh. You want OS X on x86 and they’ll want you to buy a new video card, sound card, etc.
You will not end up with a cheaper experience.
Look, come up with $1000 and get an iBook. You’ll be glad you did.
If you want to try OS X that badly, you will have to come up with the brass to get through the gate. x86 or no, you will have to buy a new computer to use OS X.
Apple, a hardware supplier, has NO incentive release a distro that you can just toss on any old x86 box.
If you don’t want to buy from Apple, try getting one of those G3 600 towers from YDL, a video card with 32mb of ram, and see if you can get OS X to install.
It’s still going to cost you at least $1200, though.
>>rather than owning a PC and bitching about it for the next 5
>I don’t bitch about my PC. It works beautifully. I bought it in 1999, and even back then was not top of the line (dual Celeron 533, 66 Mhz 256 MB SDRAM). It runs great all the 8 OSes i have there installed currently, including XP PRO.
Eugenia, having dual processor PCs is not very normal, and I’m sure the macheads will jump on that. One of my computers is a PIII-450 with 512 mb of PC-100 RAM that I bought in 1999 (ram added piece by piece over the years) have run every version of Windows since 98 (including the awful ME). I run windows XP Pro on it currently, and it works great. Startup is a little slow as it loads a lot of stuff, but from boot to full usability (less password entry) is still less than a minute and a half.
That all said, I’m still interested in playing with OS X and will probably be purchasing a secondhand mac later this year, but the whole “Macs last longer than PCs” is, like a lot of Mac mantras (PCs crash all the time, megahertz myth, etc.) leftovers from the early to mid 1990s, and do not apply to modern PCs.
>Eugenia, having dual processor PCs is not very normal,
But that does not matter, i am talking about overall speed. The fact that it is a dual, it does not mean that this is 1066 Mhz. Duals never deliver 100% of their speed, at least not on consumer-level OSes. It is more like a Celeron at 700 Mhz rather than a 1,06 GHz one. And it works great (and don’t forget that my Celerons don’t have more than 128 KB cache and they are on slow memory – my husband’s PIIIs while they have lower clockspeed are faster than my Celerons! Same goes for the Cube which has more and faster memory and a huge 1 MB L2 cache, plus it is a full G4)!
I know. I’m just trying to back up your argument by putting my even less impressive sounding PC up as an example.
One thing that I don’t get is Merrill Lynch analyst Hillmeyer’s argument that while Apple potentially could settle into a Mercedes-like market share (3-7 percent), it is not making money doing that “right now”.
OK, but the economy stinks. Apple made piles of money in the late 90s even with a small market share. And it had much weaker products — an uncertain OS transition before it, no suite of integrated iApps, no integrated mail, iCal, address book, no fast browser, no iPod. No flat-panel iMac.
And much worse retail presence. Today, there’s 85 million people
Yes, PCs are cheaper than macs (at least on desktops). But wasn’t that true in the late 90s and 2000? Wouldn’t a better economy smooth over some of the price difference for a lot of buyers?
Look, I will be the first person to admit Apple has a difficult challenge in front of it in the form of the “wintel juggernaut.” It’s continued existence is not a certaintly. But what is a certaintly in life?
I meant 85 million people within 15 miles of an Apple store.
Terrasoft Solutions is selling Power PC mother boards and computers for there YellowDog Linux OS. So all you need to do is go buy a copy of OS X and install it on Mac On Linux. Hopefully someone will port Darwin to these so you wouldn’t even need Linux to boot from.
http://www.terrasoftsolutions.com/products/boxer/
There will never be OS X on x86, ever. Not even on an x86 based closed Apple platform.
Apple doesn’t want to add support for poll() or kqueue() to OS X because they’re worried about breaking compatibility with 3rd party KEXTs. They’ve already broke binary compatibility of *EVERYTHING* once with the move to OS X. They’re not going to be switching instruction sets any time soon. They can’t afford to.
The only way OS X/x86 could’ve possibly worked is if the OS X compiler were configured to cross-compile everything to x86 as well then produce fat Mach-O binaries with both PPC and x86 support. As is there’s no application base, and therefore no reason for anyone to purchase an x86 Mac.
Furthermore, how would Apple handle a transition to x86? Kill off their entire PPC line of computers? That would be financial suicide…
Bottom line, it’s NEVER going to happen.
Hi me!!!
Look at those IP’s!
I’m not saying it’s likely, but to say “NEVER” is going too far. Here’s a possible (again, maybe not likely) future:
The G5 never arrives and the G4 falls so far behind x86 processors that a Pentium V (VI?) can emulate the latest G4 at near-native speeds. Apple builds a closed-off x86 Mac, and adds some emulation code to OS XII. You can run your old PPC apps under a transparent emulation layer at good speed, and newer apps that are compiled for x86 run blazingly fast.
You can say that’s unlikely, but they did it before. And the last time it was on a totally kludgy OS. It might be easier to pull it off again on a newer, cleaner OS.
and in the meantime I beleave this is the only reason why the are posted here. The are not really news or even os-news.
They just generate posts – like this one 😉
and raise the count for the paying customers of the banner-adds here.
I don’t bitch about my PC. It works beautifully. I bought it in 1999, and even back then was not top of the line (dual Celeron 533, 66 Mhz 256 MB SDRAM). It runs great all the 8 OSes i have there installed currently, including XP PRO.
I wasn’t directing my response directly at you, it’s the people that constantly bombard Apple news posts, moaning about OS X for X86. I’m happy you enjoy your PC it sounds like it gets the job done. I was hoping to make a point against the OS X for X86 croud not the people that are very happy with their PCs.
i knew her from the benews days
…that pretty much hits the nail on the head. All the people yanging about “go to x86!” aren’t considering how hard it would be keeping compatibility with existing Macs, and there is no way Apple would make a move that didn’t allow for that–and frankly, I don’t think that can be done in software emulation. It’s a lot easier to emulate an i686 on a PPC750 than a PPC750 on an i686 (or i786) due to differences in processor design. The signs are pretty clear that Apple will be moving to the PPC970 (“Power PC 2”?) unless something catastrophic happens.
And the only way OS X will be released on a general x86 platform is if Microsoft releases a version of Windows to run on Macs. (No, this is not a prediction.)
Even better, angryclam… I just bought a AMD XP 1800+ setup once my Abit dual 366 stopped working, and even a single proc system that much faster cannot provide the “creamy smoothness” that probably any dual setup does. Any process that hogs as much CPU power as it can (such as MPEG/Divx encoding), though it finishes much faster, pretty much locks up my system, while on the old dual I could continue to browse the web while I authored video files. The “creamy smoothness” is a quote from a review in which a user made sure to upgrade to a Tyan dual proc AMD board to continue his dual experience, like I should have done.
Eugenia, did you have your dual board modified to replace all the bad capacitors (if it was Abit)? Mine succumbed to this well-publicized problem, and I will probably get it fixed to get back my dualing.
“It’s a lot easier to emulate an i686 on a PPC750 than a PPC750 on an i686”
what is your logic behind that comment? The very fact that the PPC750 is a more elegant design should make it easier to emulate, plus the i686 has clock cycles to burn relative to the PPC750, which gives you some chance of matching its performance in emulation.
Apple have changed processor families in the past (68k to PowerPC), and they kept backward compatibility during the transition period. If any company is capable of moving from PPC to Intel it would be them.
But personally I’d rather see the PowerPC family pick up, the world needs variety, including in processor architectures, it would be a boring place otherwise.
The reason that it is hard to Emulate PPC750 on the X86 is that the PPC750 has tonnes of CPU registers, and the X86 has the common AX, BX, CX, EX. When you think about this, it would be fairly simple to just code a processor wrapper, but then again that’s a huge performance hit. You could route it through a JIT compiler, but once again we are thinking performance hit.
More information and talks about PPC on X86 solutions can be found here:
http://www.emaculation.com/ppc.shtml
It’s a mac emulation site I used to use when I owned a fairly old mac and newer PC. PPC Emulation has been promised a thousands times and never delivered, probably because there is not a lot of emphasis for it, who really needs to emulate the PPC platform to run a program on a X86 based machine? The majority (99%) of apps for Mac OS have been ported or coded to run on Windows. See there is almost no market.
i’m kinda surprised that no-one has picked up on the guardian article. in britain, the tech take-up can be quite sluggish compared to the US, and while we do have dedicated publications towards general “non windows” computing news, this is the first time i have come across an article basically asserting the positive aspect of open source software. i’m not trying to imply that the floodgates will now open and that microsoft should run to the hills, far from it, but this is a significant first step in the general awareness of other (non MS) solutions to everday computing problems.
Hi me to me too. Man, the flight from US to use a Malaysian ISP is so tiring.
It is pretty easy to write a emulator for PPC on x86. Especially when x86 nowadays have a lot of RISC features, it wouldn’t be entirely slow. It may be even faster than the fastest PPC Mac. But, like you, I think it is too early to do a transition.
The 68k transition shows how ugly a move can be, even if it was for the good. Now, the economic times are not good, while Apple haven’t fully transition to OS X, I think x86 is rather impossible. For now.
In two or three years, the story would change. But the story can change if Apple spend more money to convert Mac OS 9/< to OS X instead of Windows/UNIX to OS X.
—
appleforever: Apple made piles of money in the late 90s even with a small market share.
It lost more money than it made in the early and mid 90s.
appleforever: Wouldn’t a better economy smooth over some of the price difference for a lot of buyers?
True. But with a looming war on Iraq, and perhaps Korean War 2.0, plus the conflict betwen Taiwan and China escelating more, and the Israel and Palestine issue *still* not resolve, while Saudi Arabia funds the rise of extremist unchecked that create terrorists, and Japan still not being able to control the yen, no, there probably won’t be a better economy climate for the coming years.
Anonymous: all you need to do is go buy a copy of OS X and install it on Mac On Linux.
You actually need more memory and perhaps even CPU time for that. Plus, Terrasoft’s motherboards and machines aren’t particulary cheap either.
It may be even faster than the fastest PPC Mac.
True, but then it’s also possible that at any given time, the atoms in your body might just fall out of sync with those in the rest of the universe, and you’d be pulled straight to the center of the earth.
Neither seem very likely. For one thing, while I’m not an Apple software engineer, I do know that the PPC chips are faster at certain computations. If I *were* an OS architect, you can be damn sure I’d make sure to optimize every nuance and performance increase I could muster. Also, any given desktop PPC processor you can buy in a machine on Apple.com has more cache in L2 than the newest AMD Athlon, and an extra L3 cache (which, obviously isn’t as fast as L3, but is almost 10 times the size of the AMD’s L2, and about twice as fast as accessing memory at peak speed on the AMD’s system bus). I’m not saying a dual 1.25GHz Mac is faster than a 3GHz Athlon, I’m just saying that emulating a foreign architecture at greater speeds than an almost-as-powerful proc does natively is improbable at best.
Considering that it took 8 years for 68K emulation to reach a point where it was faster “at certain” tasks than a real 68K processor, I would think that it would take a much longer time to have even the most reliable PPC emulation. The 68K processor has a fairly basic architecture and there has been 68K emulation since 94′, since to date we do not even have anyone distributing or selling a PPC emulator I’m guessing it may take quite a few years.