Marklar

[akamai]

Probably one of the most mysterious and important rumor surrounding Apple: Marklar. I just wanted to kick off a topic about it questioning the purpose, benefits for Apple, et cetera.

Just a quick illustration of Marklar for those non-informed persons - When NextStep (OS X) was in its early stages, it was developed for m68k black hardware. At its prime it ran on 5 different platforms. In the end, Apple only kept two trees alive - PowerPC and x86. The reason that Apple kept an Intel build is to be speculated about, and this project is codenamed "Marklar." Maybe it is to secure themselves if PowerPC fails to deliver, or if Apple goes out of business they'd have the chance to rebuild the company after blowing out Marklar at the last minute. Maybe it is also a future project, or at some time a way to compete with Microsoft - I do not know, that is why this topic exists - to speculate.

wkr, Robert

[Robo]

Threads like this make me wish Apple would put OS X on x86. It'd get software developers' attention...oh well. I doubt it will ever happen...it's probably for one of the reasons you described - if they could no longer support themselves on their own niche of the world, they'd be able to bring OS X to a much wider audience.

[scratt]

Wow!

I was not aware of this. Must have been hiding under the wrong stone.

What is amazing is that if Apple did release it they would blow windows away. I really believe that. It may not happen right away but, I beleive, in a creeping fashion that would eventually destabilise Microsoft enough for a parity to exist between the two companies and maybe (and we probably don't really want this for reasons that have been discussed before) a lead for Apple. I believe a parity will come eventually anyway (remember I said that) but over a longer time frame..

So what we have to ask is why they don't...

Is there an agreement made back in the days when Apple was trying to re-establish itself... That was when Bill Gates kept popping up on big screens at Apple Expos... Is there some gentlemans agreement, or more likely a legal sword of damoclees there somewhere?

Is it simply that Apple are afraid that it would kill their hardware sales?

I for one do not believe the latter as I think the same small informed group of computer users that used Apple hardware understand why you should do so and not rely on the multitude of flavours and consequently multi-platform problems which are Windows weakness..

Perhaps that is the answer... Perhaps supporting all those different flavours of a hardware platform is fundamentally not something Apple want to do. They have experimented with that in the old 'clone wars'...
Microsoft is constantly coming unstuck because of it... Many of its OS problems and its lack of user friendly-ness is simply because of the mess of machines their operating system has to live on...

God knows... I am confused now. But what about a bootleg copy getting out!?!? What the hell would that do to the world!

It could be the ultimate virus! A virus to cure all ills!

scratt ;)

[Brad]

"Star Trek" -- "Marklar" -- "Mac OS X on x86"

It's all the same old ridiculous story. People make up the rumor that there exists some magical version of Mac OS X that runs on off-the-shelf x86 hardware despite the absolute lack of any evidence. Darwin doesn't count.

The reasons why Apple won't go this route have been hashed out hundreds of times before. I don't feel like explaining this years-old argument in detail again; so, here are the basics:

Apple depends on hardware sales for revenue. 'nuf said.

Hardware support. Think thousands of configurations versus a few dozen. From graphics cards to motherboard chipsets, Apple would be completely overwhelmed with providing drivers and, worse, support for all the no-name devices on the market.

Say goodbye to all third party software. At best, any third party software would have to be recompiled and rereleased. At worst, it would have to be completely rewritten from the ground up. Assembly and AltiVec tweaks can't just be magically converted into x86 code. Many developers would rather give up than have to rewrite all of their software for a new architecture.

Piracy. Apple's software licensing has been fairly lax because, again, Apple depends on hardware sales to generate revenue. Apple would likely have to switch to a Microsoftian method of software registration.

Microsoft already has a killer monopoly on x86 hardware. Linux has been around for ten years and people constantly harp about how it will overturn Microsoft's domination. Has it happened despite being completely open and free? Nope! Apple (especially being closed and not free) would be no different.

Don't get your hopes up. Marklar (if it even exists) is at most a very last resort. It won't happen until either the PPC platform completely dies or hardware sales completely dry up. Neither is a likely scenario any time soon.

[Brad]

BTW, Marklar was a code name mentioned two years ago in an article by eWeek. It immediately became a thing of legends despite the fact that no other legitimate source has independently made any mention of it.

So, please, remember to take anything regarding "Marklar" with a huge grain of salt.

[akamai]

Sure, Marklar may not exist, and regarding the foregoing post - it is not supposed to be news. What I do know is that 10.0 Pre ran on an x86, and it is not an undertaking to port OS X to another processor, as everything is written in ASM, C, C++, and Basic (Cocoa and Carbon only run on top). Darwin can be installed on x86, so there is a base. Over and above Darwin there are many System Apps which make Mac OS X, the OS we know. These are Aqua, Finder, iLife, et cetera. Everything which is written in C on top of Darwin (what I mean is the platform that supports Cocoa, Carbon, and everything else "Mac") can be decompiled (using very expensive software). What you will be left with is a bunch of weird code without comments and logical variable names. It will be uneditable, because it would take very long to understand. All you would need to do is recompile it on an x86. The problem with all this is driver support - but if someone would be able to overcome this, it would be likely to succeed. I have run Jaguar, Panther, and Tiger on an x86, though slow, and on an emulator called PearPC (pearpc.sf.net, pearpc.net, de.pearpc.net). But the emulation works!

[zenarcade]

MacUser has mentioned "Marklar" from time to time at least for the last 2 years. It makes sense since yellow-box was released for WindowsNT and Windows98.

i wonder how much work it would be to recompile cocoa apps to the windows platform.

[Barto]

I'm almost sure that Apple keeps a port of Mac OS X maintained. Remember all the PENTIUM|ATHLON debug code that was in common applications like DVD Player.app?

That said... it's a dead horse, especially with the G5 being such a capable processor.

Barto

[AirSluf]

XXXXX

[hmurchison]

The question has never really been "can it be done?" we've always known that it can. Software is portable and OSX is even "more" portable. The question is "should it be done?" and that right now is an emphatic "no!"

I can't even entertain the idea of Apple making a full frontal assault on X86 hardware when Apple seems reticient to even compete with MS with an Office Suite. Brad is so correct on this one. This is born from the daydreams of Mac users envisioning very cheap hardware. Sorry...lemme know when you see pigs flying.

[Satchmo]

Apple is firing on all cylinders on the iPod front, and is making a tidy profit each quarter for the past few years. And despite what Jobs says about marketshare, he'd be thrilled if Apple even garnered 10% without the help of Marklar.

[DMBand0026]

Apple will gain marketshare, and it will be without the help of Marklar, because it will NEVER happen. So if I were y'all, I'd move right along, cause there's nothing to see here.

[thuh Freak]

i wonder how much work it would be to recompile cocoa apps to the windows platform.
recompiling for the windows platform (ie: win32, .net or whatever else ms puts out; mind you, there isn't a single "windows platform", but several. and coding for one doesn't always mean your code will run on another.) would be difficult, as everything related to the gui would have to be redone. if apple had an x86 version of cocoa, it would be significantly easier, but there is still one more piece: if an application contains architecture-specific code (ie, hard coded assembly, or something else non-portable), the task would be more difficult. most apps probably dont have anything, or only small portions, architecture specific, but the bigger players probably might have some trouble.

[zenarcade]

What I meant was this . . . . there already is a Cocoa layer for Windows, it´s called YellowBox for Win98/NT.

In theory, if an application doesn´t use the latest tech, something like a window with some buttons and tables should only be a compile away.

Maybe I have got this all wrong . . .

[stoo]

Altivec's friendly programming interface encourages (and potentially significant gains) encourage more programmers to use it, making more code architecture dependent. If you can program in C you can use Altivec: no assembler required.

/I wonder when we'll start using 74xxs at work. :D

[nowayout11]

Marklar isn't necessary as long as the PPC is not abandoned. They were probably smart to go with a trickled-down Power 4, since IBM (and especially Moto) don't seem all that interested in extending the classic PPC family with their own R&D.

I think it's more of a last resort option. There aren't many other affordable and competitive alternatives to X86 at this point.

And I wouldn't be so fast to dismiss its existance. Jobs himself said it was very easy to port the OS. They "have a world of options" available to them. And there are any number of prototype products being tooled that may or may never be seen as a retail product. That's the nature of R&D.

[scratt]

Sorry to repeat myself.....

And I repeat this is entirely an academic brain fart...

But it would be stuningly interesting to see a bootleg version of this get out.
Even a partially working block of code...

Out in the wild it could be played with by people who have the time and the interest and would be quite interesting to observe...

Does everyone else think this is silly?

scratt.

[akamai]

Scratt, if I was your Joe Doe, pissed-off Apple employee who has access to the Marklar CVS and I got fired... Well... I still wouldn't publish it... You need to understand that Apple employees need to sign multiple NDAs (Non-Disclosure Agreements), undergo drug-testing, integrency testing, et cetera before they are hired. If I would publish the CVS, I would have 20 Apple lawyers and so many legal documents from the superior court of Cupertino all over me. For this reason, leaks do not happen. It really looks bad if you have to compete against Apple in a law-suit, unless the judge is an OSX-on-Intel fan :lol:.

[Brad]

Does everyone else think this is silly?
Yes, absolutely, I do indeed wholeheartedly think it is a silly and fruitless thought and that nothing good can come of it.

[scratt]

undergo drug-testing, integrency testing, et cetera :lol:.

Wow drug testing and INTEGRENCY testing!!!! I would definitely fail the first and the second sounds painful!! ;)

As I said this was just a hypothetical question. Even if I were to get hold of a copy it would only remain an interesting oddity on my hard drive... I am so anti-PC I may only ever run it in VPC anyway!!! :lol:

[TednDi]

Isn't Marklar the alien in southpark that got eaten by a Tiger?

Marklar used marklar often in a sentence.

For example, "Come in Marklar this is Marklar, we have the Marklar in our sights we are preparing to Marklar."

see:
http://images.southparkstudios.com/media/sounds/311/MARKLARS01.wav

[thuh Freak]

my blood itches.

[ast3r3x]

No guys. Seriously. It exists. I'm using it. Brad is just trying to cover it up. He is like, I don't know, like the government or something. It r0x0rz! It's 2x faster then the fastest G5. Even the mouse is faster!!!!!!!

Brad and the others are afraid of what could happen if it became public. He probably worked on part of it actually. :p ;)

I'll post a bit to prove it's real.

if(@sys != "mac"){ FuncStart(@x86_load); }

[AirSluf]

XXXXX

[akamai]

lol... Show us some more code, or basically a whole file with vars we can understand :D

[iMeowbot]

WebObjects still includes development tools for Windows. Apple need to keep Yellow Box around so they can build Project Builder and friends, but no one else can play.

[MacMadeMeDoIt]

It isn't a bad idea for Apple to retain the development of this product. With the very slow shipments, and only promises to increase speed in a month or so, it could be wise to begin building this in as part of the "rumored" x86 emulation set, or even to build a PPC emulation set inside of the Marklar. After watching the AMD Opteron basically begin to eat Intel alive, and looking at demonstrations of a one-rack supercomputer capable of over 1.4Tflops...the G5 xServe needs more than speedy shipment to compete. It needs, quad - dual core processors, along with the ability to install them in an integrated clustering solution. The xServes are very cool, don't get me wrong. But the xServe Blade would be MUCH cooler, especially if it matched the xServe RAID, along with a new xServe RAID with a true shared FC controller for dynamic multipathing and failover. xSan supports it...make it! And why not actually drop FC anyway and jump to infiniband? This is all shipping with the AMD boxes, and working quite nicely and efficiently. With the usual modifications to allow dual-core, dual-platform compatibility - this would be the only supercomputer needed.

[Brad]

It's Xserve and Xsan; please spell them right. TIA.

[stoo]

Xserve and Xsan: are they MACs? ;)

[scratt]

shh! stop it! :lol:

[Brad]

are they MACs?
http://smilies.jeeptalk.org/ups/DeNiro/fight.gif