Slashdot Log In
Apple Builds Darwin For Intel
Posted by
timothy
on Sun Mar 26, 2000 07:21 AM
from the we-can-build-you dept.
from the we-can-build-you dept.
BluesHarp writes, "Apple's lead Darwin engineer Wilfredo Sanchez announced that he successfully built and has made available all of Darwin 1.0 for both PowerPC and Intel. Does this mean that OS X for Intel would be just a recompile away?"
From Sanchez's Avadgato diary:
Neat stuff.Apparently a lot of people are under the impression that Apple isn't going to help out with reviving the Intel port of Darwin. This is false.Getting everything built fat is a big step, but a lot of work remains. The next thing is to get installation bootstrapped so we can get Darwin onto an Intel system, and then to get the kernel running, since we haven't tested the new kernel on Intel yet, and there is limited driver support for Intel PC devices. I have a high degree of confidence that most of the user-space software will work without problems, particularly since a majority of it comes from the BSD world where Intel is the primary platform, but also because we've seen it work before in Rhapsody.
This discussion has been archived.
No new comments can be posted.
Apple Builds Darwin For Intel
|
Log In/Create an Account
| Top
| 342 comments
(Spill at 50!) | Index Only
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Apple Builds Darwin For Intel (Score:3)
Jobs, being a smart man, wants to have his options open, especially after the latest problems with chip supply. There are certainly a number of people working on X86 at Apple. THAT'S THE POWER OF MACH (beside being able to run 3 layers simultaneously). But X86 is dead and both Intel and AMD know it (look at their roadmaps). PPC isn't. Motorola is not alone - IBM is part of AIM too and they're good at cranking out chips. Apparently IBM's got a dual processor on a single chip running 2 GHz. And Motorola is having fab problem - hence there's talk that they'll liscence from AMD, which has a brand new fab.
Besides this processor speed thing tends to go back and forth like a pendulum. x86 is faster, then ppc, then x86 and so forth. Probably has most to do with development cycles. Besides Intel and AMD are pushing the definition of releasing a chip (anybody gotten any sort of quantity on 1 GHz chips? - anybody expect it any time soon?). I seem to remember hardware makers bitching about Intel not delivering in time for Xmas.
As far as Apple hardware goes, it's not as expensive as at first glance. Remember, you have to throw ethernet and firewire cards into a PC to do a fair comparison (probably around $200 for cards with equivalent chip sets).
Two CPUs (Score:3)
www.xl8.com
You put two G4s on a riser card and then plug that into an xl8 Carrier ZIF 2.0
"Using a multiprocessing ready version of its patent pending CarrierZIF, XLR8 showed a dual ZIF CPU riser concept card that allows its CarrierZIF 2.0 to implement multiprocessing using standard G4 ZIF CPU daughtercards. The XLR8 MP riser is designed to also support multiprocessing ZIF daughtercards in Apple's Power Macintosh G3 and ZIF G4 systems."
Think Transparent Red-Box (Score:3)
If you've ever seen Virtual PC or one of the other PC emulators running on a mac, you can't help but be at least a little bit impressed. Now you run one of those with 128MB of ram for it... and the emulation simply flies.
Now... In some ways you could consider this to be a Red Box (The typical codename used for a mythical pc emulation box on macos's blue code base
Anywho... a Red Box... which is half decent for things you can't get native apps for... such as that silly Access database you've been using since the mid 90's.
Now Apple had a problem with the PowerPC. Namely their OS and ALL apps were written in 68k code! So they implemented a emulator at the lowest level of the system... Later version cached code, dynamically recompiled and all that juicy stuff.... And eventually native apps were released... and speed reigned... Life was good.
Then Apps got bigger. And things broke more often.
A Modern Core was required.
So the Rhapsody project was started...
Take OpenStep 4.2 (I think) which is a BSD based OS with NeXT object extensions running on x86... and port it to the PPC.
Great. But it doesn't run Mac Apps...
So make a Blue Box... its not an emulator as such... well it is... but it doesn't emulate the CPU... it emulates everything else
The point is you get native speed...
And because its running on top of a unix, it gets some nice features, like memory protection & a GOOD virtual memory system (relatively
So you just make the Classic MacOS think its running on a NICE piece of hardware with 1 gig of ram.
Add a few tweaks to the OS, so that when it goes into a delay spin it just sleeps.
And there you go. Working bluebox. But its very separate, just like Virtual PC.
Oh Well.
Enter MacOS X.
I won't go into Carbon... but carbon is cool
Basically means you can run the MacApps OUTSIDE of their little blue box
Its another API to the Unix... You have the Darwin API (BSD), You have the NeXT API (YellowBox/Cocoa) and you have the Carbon API (Cleaned up MacOS)
Cocoa is the nifty one btw.
Anywho... transparent blue box.
Basically you make the finder in the BlueBox (your desktop manager) invisible... so that the mac's windows float in MacOS X... almost like a real macos X app...
even tho they're being 'emulated' (we have to find a better term for that! Is 'Runtime Environment' it?)
Anywho...
This brings me to the next bit
Apple has all the technology required to make emulators and these Runtime Environments...
And in fact... Other companies offer VERY good PC emulators for traditional macos.
I think you would have to be insane to not believe that Connectix was currently working on a Carbonized version of Virtual PC.
And what would be really amazing is if they managed to turn Virtual PC (the Red Box) into a Transparent Red Box which is certainly possible.
This would give you your modern hardware which runs windows/macos/osx/bsd apps.
Of course, you could also run linux apps on their with sufficient development work
Oh yeah, Java too.
In fact, you might actually be able to run LinuxPPC as a separate Mach Process!
BTW, Aqua is even better in person!
---
Live Long & Prosper \\//_
CYA STUX =`B^) 'da Captain,
Re:AltiVec (Score:3)
This is not the case. Apple does not currently sell any dual-processor machines, and has not since the days of the PPC 604e-based Power Macintosh 9x00, which ended in late 1997 or so.
The rationale for this is that the current Mac OS takes ill advantage of multiprocessors. It uses assymmetric multiprocessing, and the second processor sits idle for the vast majority of time. (Furthermore, the G3's 3-state cache coherency model is insufficient to support more than dual-processor configurations, although the G4's 5-state model is enough to support it very well.) Mac OS X will fully support symmetric multiprocessing, so expect this picture to change in the next year or so.
Re:Apple's profit margin (Score:3)
I think you've seriously misclassified Wilfredo Sanchez. His job is managing Darwin and Apple's other recent public source projects. He (and now David Zarzycki) are the conduits through which Apple and the open source community communicate. He's quite dedicated, and this is a big step; cross-compiling an entire operating system is far from simply "a [single] bored engineer blowing off steam." This build required that Apple's core OS development team be fairly rigorous in writing and maintaining cross-platform code. (Admittedly, Darwin is itself a port of OpenStep/Mach's core OS to PowerPC from a codebase that ran on x86, SPARC, and PA-RISC, but many, many, many changes have been made.) Not only that, but the build infrastructure is now such that Apple's public and private CVS servers are largely unified for the core OS; we're getting a live view of of Mac OS X's core OS' development.
Meanwhile, the modified GCC which Apple inherited from NeXT has an engineer dedicated to merging the codebase into the mainstream GNU GCC, copyright reassignment and all. There is even serious discussion of bringing together the GNU and Apple Objective C runtimes.
As I've said in other posts, I don't specifically do not believe Mac OS X will be running on Intel at any point in the forseeable future, but it is my opinion that you've sorely misjudged the rest of the situation.
Re:Yes, but none of that matters (Score:3)
If Apple can maintain 5-10%+ marketshare, then they'll be around for some time. If they can hit 15%, then things will be perfect. You don't have to have a monopoly to survive.
(BTW: People have been anticipating Apple's death for 25 years - I would hope that they'd learn by now that it's just not happening)
- Jeff A. Campbell
- VelociNews (http://www.velocinews.com [velocinews.com])
Re:i think there is good in this. (Score:3)
There's no point in that really. OSX uses a seperate backend display system to draw everything to the screen, this is really the only thing where most of the work needs to be done to run with a different graphics environment. There is already a yellow box for windows that allows Cocoa apps to run on win, I think you have to buy WebObjects to get it though, apple hasn't really decided what to do with it.
For example, the variations in GUI interface from product to product/release to release -- one has gnome and E, another KDE, and the newer stuff will have sawmill and gnome...these larger companies don't want to drop big changes on the employees whenever some distro changes their desktop functionality. The mac and Win GUIs are far more stable than the free/open systems.
I definately agree. I think the GUI on linux is really going in a bad direction in that there are too many parallel efforts. The desktop environment is really where linux sucks right now. Yeah, I said it. But, consider cut and paste. Cross-application cut and paste is pretty weak in linux. Sure, you've got the middle mouse button, but to paste a url from a terminal to netscape means you
- selected the test in the term
- bring up netscape and click the middle button at the end of the url
- delete the old url
This is not ideal. And what about cut and paste, and universal helper applications? These are things that Gnome and KDE are addressing, but how compatable are we talking? What if I want to run WindowMaker? Will I only get the benefits of gnome's app integration with gnome apps, and kde's with kde, etc?
What we really need, IMHO, is a standard for system wide application integration and communication rather than several competing efforts.
BeOS? It's not being used by enough people to be taken seriously. And the last time I looked at their developer area, I think I saw they lean toward objective C. Regardless of the qualities of the language, PHBs want to see C++ prominently displayed in the developer area, like it or not. They don't want to retrain the less gifted developers.
Well, there is Obj-C++ (shudder)... Obj-C is a great language and I really hope apple can begin to better market it rather than turning to Java. I know a lot of people don't take OpenStep stuff into consideration because of this crazy Obj-C thing, hopefully that will change...
The final issue is the toughest to face -- price. No one will like to hear this, but for OSX to rapidly gain acceptance, Apple will need to sell a $19.99 unsupported version. It's the only way to load the home and corporate desktop with this OS. Even the full version can't sell for $199. That's more than Win2K. And they can't sell it for $99; Linux and BSD will beat it to death as their desktops stabilize. They can't make it by selling it as a server OS for the back office only. That philosophy will lead to nothing.
I think more important is for them to beef up developer support and keep the dev tools free (interface builder and project builder). Interface builder is by far the best RAD GUI develpment app in existance. Project Builder is pretty cool, but supposedly there will be a beefed up version for OSX which will improve some of it's weaknesses. Anyway, I think the key to the success of OSX is getting developers interested in Cocoa and the OpenStep way of doing things. Microsoft got ahead because of great developer support, apple needs to do this now. Not charging for the tools and updating some of the old OpenStep docs will do a lot I think.
I don't think sales cost is really going to be a big deal at all. It will probably be under $100 as past versions were, and hopefully a good upgrade price. I really doubt a super low cost is going to help them make it much more of a success. Attracting developers is going to be the big thing. The more developers they have, the more apps they have, the more people will want to use OSX with advanced new apps, the more computers they sell... People aren't going to want to run out of date apps in the BlueBox forever, and I don't know how long anyone is going to want to maintain carbon apps...
Re:Apple's profit margin (Score:3)
Darwin is an OS. Go check it out. It's not just a kernel. The only major thing missing from it is a windowing system. It includes a kernel, shells, compilers, drivers, servers, editors, etc etc. Please at least look into things before making uninformed statements.
a percent of the code? No, darwin includes most of the lower levels of OSX. The things it doesn't include consist mainly of the window system, Cocoa and it's frameworks, the bluebox, Carbon, and Aqua. Of course John Carmack has devoted himself to porting X to darwin with it already running on OSX Server, so I suspect it won't be long before we have a window system for it.
A problem though... (Score:3)
The gcc port for PPC pretty much blows. If you do some benchmark under MacOS vs Windows on Intel, the mac will kick butt for the same processor speed (as it can do more per clock cycle). However, if you do something like run nbench on LinuxPPC vs Linux on Intel, the intel will probalby pull ahead because the compiler doesn't do optimizations very well for PPC. It kinda bums me out that I have this nice fast Mac but there's no real advantage using linux on it because the performance boost from the hardware just isn't there.
Hopefully some of apple's changes being incorperated to gcc will fix this, but I doubt it. The compiler on OSX Server isn't so hot either. Of course, it is based on a really old gcc, so I suppose it's possible that they've been tweaking up a moder and super fast gcc for OSX. one can only hope...
darwin is not everything (Score:3)
it was a cross-compile, not a native build (Score:3)
Darwin was cross-compiled, using a PPC, into binaries for both x86 and PPC.
Apple's profit margin (Score:3)
This foothold is limited to Quicktime and a few other proprietary formats that, while wonderful in their own right, have a limited lifetime (as everything in the computer industry does).
Were apple to suddenly decide that it were to be quite profitable on their part to sell OS X as x86 compiled code, they might get a larger install base of software. However, they would not do this because they may lose business on the hardware front.
Think about it: Apple's OS X is going to be a Big Thing. It is going to be hyped up the arse. The only way to use it will be to run on Apple hardware. Apple sells machines to the unwashed masses so that they can use the over-hyped software (conveniently also provided by Apple.) Apple makes a killing.
Where is the advantage of selling the software? Perhaps if they implemented a licensing scheme like Microsoft, or perhaps a help scheme like RedHat (which would be a terrible idea because Macs are so bloody easy to use).
I'm terribly sorry to say (I love apple, and I would love to see OS X running on my Intel box) but this is just another case of a bored engineer blowing off a little steam.
Rami James
Pixel Pusher
Altec Lansing R&D, IL
--
AltiVec (Score:3)
I took it for granted that Mac OS X would be utilizing Altivec's special capabilities to make some significant speed boosts. When recompiling for x86 compatible machines, wouldn't those speed boosts be lost? And therefore wouldn't the version that runs on Apple hardware be significantly faster?
Just curious; I don't really know that much about Altivec and Apple's coding practices..
Rami James
Pixel Pusher
Altec Lansing R&D, IL
--
Re:AltiVec (Score:3)
In brute processing terms, from the benchmarks that I have personally witnessed, the G4 wins out everytime.
The main performance hindrance for Macs has been the actual Operating System. Hopefully with that out of the way, we will have some serious workstation power within the price range of normal human beings (a 1000Mhz Intel/AMD chip is not within the price range of normal human beings.. more like pod people if you ask me).
Rami James
Pixel Pusher
Altec Lansing R&D, IL
--
Re:Apple's profit margin (Score:3)
Would you rather have a car that gets you where you are going, or one that will get you there looking good? Would you rather wear a pair of 10 dollar sneakers or those nice, cushy ones that you paid 80 dollars for? We live in a society that has an economy that makes it so that we pay more if we want more quality.
Have you ever cut your fingers up while trying to replace a stubborn CD-ROM drive in a poorly designed computer housing? Try doing that in a G4 case.. you'll never want to go back to evil PC cases again.
More quality == more money.
Rami James
Pixel Pusher
Altec Lansing R&D, IL
--
Re:Hello? It IS broken! (Score:3)
I saw a demo of OS X just a few days ago. I must say that I was stunned at it's features. I guess what I quoted above is partially correct. If you listen to everyone talking about Linux, then there will be a lot more people. But how many of them actually get it? Most of them are just going along with it because every one else is. If they saw the capabilities of OS X, they'd be singing a different tune rather quickly.
I have been a PC guy, but after seeing OS X in action, I have to say that it really is a powerful system and should not be overlooked just because you don't like Macs.
John Lavoie
Ithaca College
Whats next after Darwin? (Score:3)
what's wrong with this picture. (Score:3)
last time I checked, software was much more profitable than hardware since it costs so much less to produce and distribute. if you sell s/w that's not tied to a hardware, you sell your stuff to anyone and are not limited by the production limitations found in the h/w business. plus your margins are much higher if you sell enough (after covering development)
think about this, if they sell OSX only for apple h/w owners, and they are the sole providers of apple boxes, then they'll sell as much s/w as they can build boxes, which have a completely different business model (based on low volume, high margins vs. high volume decent margings).
after all microsoft made its money selling software when the internet wasn't even around and computer usage was just a fraction of what it is today, but it took a revolution, new economy, new age, the internet, etc. for cisco to have as much money as they do today sellin hardware.
IMHO: keep selling apple h/w adn s/w turnkey solutions, AND sell OSX for Intel for those who want it. whoever was going to buy an apple box will still probably buy it, the rest of us that would like to give it a shot but would never spend on their hardware (i love my custom built PCs) will probably shell out $50 for their s/w.
one more thing, you have OSX on intel and a very good, reliable, fast win emulator will show up in a week. THEN you really start getting sales in by having the still missing critical mass of s/w apps for your OS.
Of course not. (Score:3)
Re:AltiVec (Score:4)
Re:Whats next after Darwin? (Score:4)
It seems lately those guys have been more interested by the embedded market for the PowerPC line of processor than for the desktop market.
The G4 is still stalling at 500 MHz, that's more than 6 months after its introduction... meanwhile Intel and AMD are delivering over 1 GHz processors.
Maybe Apple is considering switching to x86 processors as a back-up plans if things gets really bad with the PowerPC for the desktop.
For those who think that wouldn't work because People wouldn't buy Apple boxes anymore and just run MacOS X on whatevet cheap PC hardware is available... think again:
a) nothing prevent them to adapt their OS pricing scheme to be profitable from mass OS sales
b) they will still provide the best plug & play experience since they will make sure MacOS X works smoothly with all their hardware
c) all the usual Apple stuff some of us love (Firewire, clean design, nice cases, inexpensive wireless, etc.) will probably keep the mac faithful to buy Apple hardware
Honnestly, those days, beside for the processor and motherboard, the parts are usually the same in a Mac and PC (IDE, USB, PCI, AGP, etc)... and the processor is on a daughtercard...
Janus
What's next after Darwin... (Score:5)