Slashback: Alternatives, Ads, Apple

Slashback tonight with word on what you can use instead of JPEG, the return of AdCritic.com as payware, NCR vs. Handspring, and more. Read on for the details.

Prepare the JPEG bonfire. Moderator writes: "Here is an open alternative to the JPEG file format. I tried posting it in the JPEG patent article but it got buried under all the comments about "THEY CAN'T DO THIS!" and stuff."

This project is called DjVuLibre and encompasses "a set of compression technologies, a file format, and a software platform for the delivery over the Web of digital documents, scanned documents, and high resolution images."

I hope the judge has a big "WITHOUT MERIT" stamp. theodp writes "A U.S. District Court has issued a summary judgement in the patent infringement lawsuit filed against Palm and Handspring by NCR, dismissing NCR's suit as having no merit. Praising the decision, Handspring's CEO said 'Settlement of this case was never an option,' while Palm's CEO remarked 'We refuse to succumb to intimidation by companies that use charges of patent infringement to bully others.' One of the NCR patents in question was for 'a portable terminal small enough to fit in the user's hand,' and the complaint went on to claim that NCR's researchers, 'recognized an unsatisfied need for a portable, handheld device which would allow the user to information such as appointments, to-do lists, and addresses, and execute financial and shopping transactions by connecting to networks using an interface module.'"

This is sure to bring out the AdCritic critics. thebus writes: "The good news. AdCritic is Alive! The bad news. You gotta pay!"

An annual subscription for $69.95 looks like something worth paying for if you're in the advertising industry, but it would be nice to get a less expensive "interested viewer" option as well. Oh well.

Oh Steve, ya big tease! Maïdjeurtam writes: "In this Yahoo finance article, Reuters asked Apple's CEO Steve Jobs about the possible abandonment by Apple of Motorola and IBM's processors (PowerPC G3's & G4's), and the possibility of Intel processor-equipped Macs. Steve Jobs didn't exclude the possibility. He noticed that, during the year 2002, Apple had to finish the OS X transition and, this done, there would be a lot of amazing possibilities, which he finds exciting."

Most of the content of this article was covered in yesterday's coverage of Jobs' keynote, and the bit at the end about other processors may be only a throwaway line, but it certainly is intriguing.

Comments

A jpeg replacement

[Anonymous Coward]

How many of you are using the GIF replacement PNG? Anyone... anyone? Yeah, exactly. Unless IE (for Windows) supports it right, no one will ever use it for the web. THAT is why we don't want anyone company to develop a monopoly.

PNG in Mozilla (and Opera) is pretty darn great. And 24-bit transparency rocks. Too bad I can't use it too often.

Re:A jpeg replacement

[dattaway]

I'm thinking about using png exclusively. If the content is good enough, I'm sure many IE users would make the effort to swtich.

Re:A jpeg replacement

[bedessen]

I'm thinking about using png exclusively. If the content is good enough, I'm sure many IE users would make the effort to swtich.

Why would they need to switch? IE has supported png since version 4.

Okay, so some features aren't implemented right and there are still a few gotchas, most notably alpha channel support. Surprisingly, IE for the Mac has perfect alpha channel rendering, yet it remains broken on the Windows version.

Re:A jpeg replacement

[James Foster]

FYI, Internet Explorer does support PNG.

Re:A jpeg replacement

[billatq]

Actually, I know of a hack that someone came up with that allows internet explorer to properly use transparent pngs. Look and be amazed! Real browsers ignore the microsoft crap and render the png properly, and IE uses a directx filter to make the png transparent. See http://www.tidakada.com/ to see it in use.

Re:A jpeg replacement

[L. J. Beauregard]

Don't the patents on GIF89a run out early next year?

The Unisys patent expires June 30, 2003, 20 years from filing.

Re:DjVu libre link

[Pxtl]

I've never even heard of this format before. Anyone know how far its gotten (plugins for major apps, built-in support in anything, a single M$ app that you can use it with)?

Re:DjVu libre link

[God! Awful]

A link to DjVu libre wouldn't have hurt. Here it is: http://djvu.sourceforge.net/ [sourceforge.net]

What are you talking about? It's right there in the story submission.

-a

DjVu not an option

[Guspaz]

It's *nix only, which makes it useless to the vast majority of users. Sure, you can get a plugin for Windows browsers here http://www.lizardtech.com/download/?f=0&d=1 but that doesn't mean you can compress images.

A more suitable alternative is JPEG2000. And if this patent thing helps it's rollout get along faster, I'm all for it.

Regards, Guspaz.

Re:DjVu not an option

[King of the World]

JPEG200 is more encumbered by patents than GIF. It'll never see the light of day.

Re:DjVu not an option

[jkramar]

>It's *nix only
That isn't fully correct. LizardTech provides encoders for Windows, although at a price. You're right that it doesn't have much support (notably a free Windows encoder), but JPEG2000 isn't widely supported either; no new format ever is. DjVu could go somewhere, although just where cannot be known yet.

Re:DjVu not an option

[hyperturbopete]

Well, there was the GIF patent and it took what, around 5 years to get decent PNG support in most apps? (but it finally works well in web browsers, office progs, photoshoppish progs, etc)

If JPEG goes the way of gif, perhaps people will realize its in their benefit to stay w/ and open format, and the next major release of everything will have whatever the "open" alternative is... but realistically its probably gonna take another 5 years
_

Re:DjVu not an option

[Doomdark]

Plus the names awful, DjVu (deja vu) is almost as awkward as Ogg Vorbis. Why must so many open source alternatives have awful names?

And what exactly is cool about acronyms like gif, jpeg or mp3? Nothing except that many people recognize them. They are not popular because of their name, they are popular since they are popular. And to become popular, they needed to become ubiquitous standards.

Re:DjVu not an option

[Doomdark]

Heh. The other meaning for gay might lead to file suffixes from the 60s Batman series. Then we'd be swapping "paf", "kaboom" and such files?

I must admit that suffixes like "gif" and "zip" are easier than, say, mpeg or jpeg or mp3, being mono-syllablic. In that way "ogg" is not too bad.

Re:DjVu not an option

[Tet]

It's *nix only, which makes it useless to the vast majority of users.

Errr, no. DjVu is a compression format, and hence inherently platform neutral. In fact, the windows encoder predates the Unix one, IIRC. Either way, there's a free reference library, so it's easy to incorporate DjVu encoders into a Windows art package, for example.

I hope Apple keeps Motorola

[Knife_Edge]

I like the idea of a significant number of processors in the PC market that are not made by the two chip giants AMD and Intel. This stems partially from concerns about things like the pentium's processor serial number now - and future possibilities like palladium. I also appreciate that, thanks to the G4, Intel can no longer claim clock speeds are the only meaningful chip performance measurement.

Re:I hope Apple keeps Motorola

[Resist148]

I wouldn't hold your breathe for Motorola, not that I don't want them to be successful, but they just reported a loss of 2.3 billion dollars last quarter. That makes 6 quarters in a row where they have lost money. They just can't go on much longer losing that much money. Maybe if they can get a new Processor out the door....3 year old 800mhz processors aren't going to cut it too much longer.

Re:I hope Apple keeps Motorola

[supabeast!]

That's cool and all, but some of us like the idea of cheap CPUs. Remember how overpriced Intel chips were before the K6-2? And how slowly intel ramped up CPU speeds? Competition in the X86 world was a massive boon to CPU consumers, just like competition by the Dramurai was for RAM buyers. Perhaps if Apple started offering X86 and Motorola options, Motorola would get it the hell in gear and sell faster, cheaper stuff.

Re:I hope Apple keeps Motorola

[MoneyT]

They could still stick with IBM. But your right, I haven't liked motorola in a long time (they held up the G4s) and they've been loosing too much ground to stay viable.

Re:I hope Apple keeps Motorola

[Christopher Thomas]

I wouldn't hold your breathe for Motorola, not that I don't want them to be successful, but they just reported a loss of 2.3 billion dollars last quarter. That makes 6 quarters in a row where they have lost money. They just can't go on much longer losing that much money. Maybe if they can get a new Processor out the door....3 year old 800mhz processors aren't going to cut it too much longer.

Since when is Motorola's sole product line PPC chips? Motorola manufactures a vast range of semiconductor and consumer electronics products.

Whatever the source of Motorola's woes, I doubt it's the PPC.

Also, it's only the iMac that has an 800 MHz chip. The PowerMac G4 has 1 GHz chips. I'm sure there will be speed steppings down the road if Apple feels threatened by Intel and asks for them.

Re:I hope Apple keeps Motorola

[anothy]

god, not the IA-64. lets keep that confined to the windows ghetto as much as possible. it's about the worst chip design i've seen, ever. AMD's 64-bit design is way better. but even there, i'm not clear why they'd go with that vs. the PowerPC chips (64-bit versions exist, they know PowerPC works for them, and have no conversion) or moving to IBM's POWER4 (64-bit, proven history, similar to PowerPC for a simpler transition). if Apple's going anywhere off PowerPC, i'm betting on IBM.

Re:I hope Apple keeps Motorola

[anothy]

Intel chips are still overpriced for what you're getting.
anyway, the suggestion of supporting both is interesting, but has proven several times to be commercially impractical. the most obvious example being WinNT. the most popular platform becomes the focus, the others are neglected. the focus for the OS and systems designers, but - much more importantly - the focus for the app designers. WinNT for MIPS and PowerPC were canned primarially because there were virtually no apps for them.
mind you, OS X is actually in a much better place for this than NT was/is. OS X has some nice features for multi-platform binaries, and (assuming Apple is still building Macs, rather than running on commodity hardware, which is all but impossable) the difference won't start out nearly as great as they were between PCs and Alpha+MIPS+PowerPC workstations.
but even so, it doubles the amount of testing and integration work for the app writers, and doubles pretty much the entire body of work for the OS writers. two-platform binaries would be almost twice as large as their specialized cousins, and may or may not come with a startup performance hit (depending on how OS X implements this feature - i don't know).

and, most importantly, for what gain? how much does the use of a PowerPC processor really add to the total price of the system? i can't find prices right now, but i can't imagine it being much more than ~$100. is that really worth it to Apple, or their customers? not likely.

Re:I hope Apple keeps Motorola

[anothy]

uh, what?

IA-32 is a pretty poor design, compared to most other chip designes currently on the market. IA-64 maintains all its crap, adds HP legacy support (basically both as seperate cores), and adds the new IA-64 VLIW/EPIC instructions on top of that. VLIW's never seen a commercially succesfull implementation, nor one that performs anywhere close to more traditional processor designs. nobody's ever thrown as much time or money at it as Intel is, so it'll be interesting to see if they can keep it competative, but that doesn't change the design issues. and EPIC is a particularly risky (although interesting, from a research point of view) implementation of VLIW. the whole thing's a mess, from top to bottom. and that's before even talking about what the design's going to force on compiler writers. ick.

anything would be better than that, but AMD's 64-bit chip is actually not to shabby on its own right. it still keeps the IA-32 legacy crap, which i think is bad (although understandable), but is much better in nearly all other respects. it doesn't improve the 32-bit portion, no, but all the 64-bit stuff is done much better.

mind you, i'd still rather have an Alpha, Sparc, MIPS, PowerPC, or POWER chip (in no particular order) inside my system.

Re:I hope Apple keeps Motorola

[Spyky]

The POWER4 instruction set is *identical* to a PowerPC instruction set. It is lacking only the Altivec "multimedia" extensions of the Motorola G4. However, if IBM were to produce a scaled down version suitable for desktop processors, it would seem possible to add the extensions to its instruction set. That said, the POWER4 is a fantastic processor mostly because of its huge memory and cache bandwidth, huge interprocessor communication bus, and flat memory model. It is intended to support large-scale multiprocessing. It would be necessary to scale down the cache and memory to make the processor affordable enough for a desktop machine. Doing so would kill much of the performance advantages of the POWER4 (okay it still has kick ass branch prediction, rivaled only by the Alpha).

At any rate, I don't think the POWER4 is really appropriate for a desktop architecture. It would require pretty significant modifications to make it suitable for a desktop processor (ie, cheap enough). Not that IBM couldn't do it, but it wouldn't be a POWER4 anymore. Not to mention that it might be hard for IBMs sales division to explain why processors of the same name appear in $2k desktops and $500k servers.

After saying all that, I really want to see someone produce a dual CPU on a single chip, and I want my next Mac to have one :-) or maybe 2. I think IBM is the one to do it.

-Spyky

Re:I hope Apple keeps Motorola

[Spyky]

The IA-64 is an entirely new architecture, it has no support for IA-32 instructions. IA-32 architectures would have to be emulated to run on the IA-64. There may be some plan to packing an IA-64 architecture processor with another core that runs IA-32 natively, but I am not aware of any such design. To say that IA-64 maintains all of the crap of IA-32 is simply not true.

It is also not true that no one has produced a commercially successful VLIW machine, the processor in the PlayStation 2 is an example.

Have you actually looked at the IA-64 architecture? It is astounding. Complex certainly, but astounding. Predication registers are very impressive, but require ISA support, something that will never be processor in AMD's legacy extension architecture.

In deciding between IA-64 and AMD Hammer, the IA-64 wins. Even AMD knows that the Hammer is simply a transition product. They are hoping to steal the market from Intel by producing a "64-bit" chip sooner that Intel.

-Spyky

Re:I hope Apple keeps Motorola

[Lars T.]

Yeah, but how long is AMD going to be there? They are also bleeding. And if they are gone, even Motorola will again be faster than Intel.

Re:I hope Apple keeps Motorola

[MountainLogic]

It would be an interesting business model for Apple to release X for x86 (would that make it x86X). It would be a new income stream for them, and cut directly against MS's core cash cow. The real question is would it hurt their hardware business? I think for the near term loyal mac users will stay with "known good" hardware. It would be interesting to see a OS price war.

Re:I hope Apple keeps Motorola

[anothy]

//The POWER4 instruction set is *identical* to a PowerPC instruction set.

are you sure? it this is true, i'd be quite impressed (and pleasantly suprised). i know the original PowerPC arch started as a fork of the POWER development, simplifying the instruction set significantly (and helping IBM retain much of its flexability for future development). it'd be pretty impressive if IBM has merged its POWER architecture with the PowerPC one. can you provide some reference for this?

note also that (i think) the altivec stuff isn't even technically part of the PowerPC instruction set, but is defined as an extention.

Re:I hope Apple keeps Motorola

[anothy]

...you've either not actually looked at either...

nope, i've looked at both.

...and are basing your viewpoint entirely from what you've read people posting on /.,

if i though for a second that were true, i'd turn off my computer forever.

...or you really don't have a clue about chip design.

well, i fully admit to not being an electrical engineer, and not being an expert on any chip design, but i certainly "have a clue".

i'll agree fully that IA-64 is provocative, unusual, and arguably innovative. but none of those make it good. i'm aware of the unusual design decisions Intel's made with IA-64, but you have to at least concede the complexity of the design is way above AMD's 64-bit design (or pretty much every other 64-bit design). the compatability with multiple previous existing instruction sets is killer, for one. and the added burden their model will put on compiler writers (to get even reasonable performance) is tremendous. and for speculative gains - no VLIW design's ever performed like the theorists hope. AMD's design is more conservative, to be sure, and less "exciting", but much easier to write for, much simpler. what is it you find "crappy" about it?

Re:I hope Apple keeps Motorola

[anothy]

...for the near term, loyal mac users will stay with "known good" hardware.

maybe. but that pretty much sacrifices any chance of growing their hardware/systems market.
also, for X-386 to cut against MS's cash flow would require Apple to cut into the bundled licensing deals with people like Gateway and Dell - and we all know how that goes. MS exerts significant preasure to prevent even dual-booting scenarios (note the Be lawsuit), and to cut into MS's revenue stream, it'd have to be single boot, with no Windows. simply not gonna happen.
also, as you suggest, the sustainability of Apple's hardware business in such a new model is only "near term". it's highly unlikely that it'd last against the marketing powers of folks like Gateway and Dell, if they start pushing OS X compatable machines. the vast majority of people will take the $200 price cut for the generic system.
i've got a few x86 boxes i'd simply love to run OS X on. it's a beautiful (aesthetically and from an engineering view) system. but i strongly hope Apple stays well away from that market, for their own good. at least for several years yet.

Re:I hope Apple keeps Motorola

[anothy]

yes, i've looked at both Intel's and AMD's 64-bit architectures, as well as being familiar with several other 64-bit architectures.

The IA-64 is an entirely new architecture, it has no support for IA-32 instructions.

ah, good catch; i was being unclear. yes, the instruction sets are different. but all the IA-64 chips announced to date retain 386 compatability. that includes everything from MMX through Virtual 8086 and Real mode. ick. and because the IA-64 arch doesn't handle IA-32, as you note, the handling of compatability becomes very hairy. switches between modes are non-trivial, and duplicates most of the work of booting a '386. IA-64 doesn't maintain the IA-32 crap: Itanium and friends do.

...the processor in the PlayStation 2 is an example [of a commercially successful VLIW machine]

hmm. MIPS cores, while arguably VLIW, are much closer to traditional RISC machines than Intel's EPIC, for sure. i'd also note that the MIPS chip in the PS2 has a much more limited function than what Intel's aiming at. VLIW's a murky term, and i wasn't including MIPS chips, although that's a reasonable thing to do, in which case, yes, i stand corrected.

AMD's design is more conservative, less innovative, and less exciting. the handling of 32-bit code complicates the instruction set over Intel's 64-bit-only model, but dramatically simplifies the resulting chips. AMD's design is also more predictable. it's more familiar, and dramatically less work to write for - particularly to get anything resembling reasonable performance.

all this being said, i hope to be proven wrong. i'd love to kiss IA-32 binaries goodbye, and find Intel (or someone) producing an IA-64-only chip with no support for the legacy crud. but i'm just not convinced that IA-64 is what's going to make this happen. Intel's taken too many chances at once, and pushed too much of the problem onto the software writer's shoulders.

the fact that there's so much wrong with the chips (and cores) comming from both AMD and Intel just re-affirms by stance that Apple - and everyone else - should avoid that whole mess. PowerPC, Alpha, SPARC, MIPS, and POWER all have 64-bit down better than either of the upcomming solutions.

OSX on x86

[JanneM]

Of course, _nothing_ says that any apple x86 computers would in any way be compatible with standard PC offerings. They would likely still have their own, special BIOS and architecture, and would likely include some 'special', cool, apple-specific hardware OSX would depend on. You would not be able to get OSX to run on anything but genuine Apple hardware, x86 or not.

/Janne

Re:Good god please!

[tupps]

My guess that while it will be running an intel processor it won't run on 'Commodity' PC hardware. The fact that Apple controls the hardware and the software is what makes a Mac so great.

Thanks

Luke

Re:Good god please!

[gwernol]

Am I the only one who thinks that this should be an *entirely* obvious step that should be taken immediately?

I'm sure there are many users who share your wish. Sadly it is hard to see how this makes sense for Apple. They make their money from selling hardware. It is conceiveable they could transition to making their money from selling operating systems - though only Microsoft has ever really built a significant revenue stream in this business. But managing the transition would be almost impossibly hard.

It is almost unimaginable that a public company could turn to its shareholders and say: we have a sustainable business that makes about $1 billion a year and isn't under threat. We're going to throw that away and go into this other business, with Microsoft as our main competitor, and we might be profitable again in a decade.

It just isn't going to happen.

What Apple will do, IMHO, is start selling Macs with Intel-compatible CPUs in them. Mac OS X would only run on those Apple machines, not on an off-the-shelf PC box.

This makes perfect sense, given that Intel's economies of scale and the competition from AMD means their CPUs are faster and cheaper than PowerPC. Apple can still make its money off hardware and the value of its industrial design.

Anybody that pays for advertising ....

[SirSlud]

.. is a sucker. Sorry, had to say it.

I know, some are really really really funny, but sometimes one has to make a stand for one's pricipals.

I know hollywood movies can and have been one huge ad before (Wizard comes to mind for Mario 3, Pokemon, Big Trouble was a massive dorito ad), but doesn't anybody take issue with the fact that music and movies for pay hasn't come about yet, but that advertising for pay might? Isn't that kind of twisted?

Re:Anybody that pays for advertising ....

[SirSlud]

>Dave Barry did mention the fritos... quite frecuently...

Oh I believe that, I just call it an ad because it functions as an ad. (Especially the references to it being the food of choice for the jesus character.) There is a decreasing inventory of cultural references that are not related to commercial brands, and I find it fairly depressing considering the fragility of the brands - unlike non-commercial culture, they can disappear off the radar for reasons completely unrelated to social and cultural events .. private interests affecting what might be argued as the last bastion of public space.

I work in advertising too; only in programming capacity. I can understand why it would definately be useful for you. I'll check it out, I may have spoken too soon - perhaps they are not aiming their service towards the user that wants to see that funny ad 20 times. It can certainly be useful for the agencies ...

Re:Anybody that pays for advertising ....

[SirSlud]

I thought I paid for cable, and my bill was partially subsidized by advertising ..

Expound, please. :) Advertisers pay money to get their commercials aired, so why are they paying for it if I am?

Re:Anybody that pays for advertising ....

[artemis67]

.. is a sucker. Sorry, had to say it.

Well, I work for an advertising agency. AdCritic is a very useful tool and worth paying for. There's a lot of commercials that don't get nationwide airplay, and there's a lot of stuff you'd want to watch when you're concepting for a client's tv spot.

Most of the marketing managers at medium to large sized organizations would likely find it as a useful tool, not only to keep up on trends in tv advertising, but also to keep tabs on the competition.

And I'm sure that there are marketing students in undergraduate and graduate programs nationwide that might find it a useful tool in doing research.

I bet a lot of directors, producers, film editors, etc., would also find it a valuable service.

Re:Anybody that pays for advertising ....

[Enigma2175]

Well, I work for an advertising agency. AdCritic is a very useful tool and worth paying for. There's a lot of commercials that don't get nationwide airplay, and there's a lot of stuff you'd want to watch when you're concepting

You didn't even need to mention you were in advertising. Once you used a noun as a verb ('concepting') it was blatantly obvious.

Re:Anybody that pays for advertising ....

[greygent]

I didn't like my principals in school, they always gave me detention.

Sometimes, I make a stand for my principles, though.

Porting OS X

[dhovis]

I think that this is one of the reasons that Apple maintains an Intel port of Darwin. It is not Apple's intention to move, but as long as the code works on more than one architechture, it makes it easier to port it to another one. You will loose Classic support though, and everything will have to be recompiled, but MacOS X provides a mechanism for distributing multiarchitecture binaries.

IA-64 seems more likely than IA-32 to me, but some people have suggested Apple could move to IBM's Power4 line, which is closely related to PPC. How about ARM Processors? MIPS? Sparc? Alpha? Transmeta? Anybody have any other ideas?

Re:Porting OS X

[Skuld-Chan]

Anybody have any other ideas?

MC68K!

Re:Porting OS X

[binaryDigit]

Power4 - Won't happen. The "chip" is too high end to attract enough customers. Perhaps IBM will come out with a new PPC chip based on Power4 concepts (multiple cores, high bandwidth interconnects) though.

ARM - Probably won't happen. ARM is pitched toward low end PDA class machines, OSX would be way to heavy for such a use. Perhaps if it were paired down and classic compatability thrown away, it _might_ happen. Though it's hard to imagine any of the current ARM based PDA guys switching away from M$ since they all have desktops that use Windoze and they wouldn't want to piss M$ off.

MIPS - A moribund processor (as far as the desktop market goes). Not enough market share and no real hope of ever getting any (unless the PS2 with Linux kit starts to take over the home pc market space ;) Maybe Apple will buy SGI and whatever patents they have remaining?

Sparc - Not likely. Those who buy Sparc based machines are typically data center guys, who would laugh in your face if you suggested installed OSX on their E10000.

Transmeta - Now this might have been interesting. Tweak the chip to efficiently run Classic as well as native OSX. Not enough ooomph though. Apples biggest marketing headache against WinTel is the mhz wars, and TM would not help there at all. And of course TM's problems. Perhaps IBM will buy TM and incorporate the technology into the G6 (along with the Power4 stuff from above).

Alpha - Never. A dead end as far as the market goes. Though running your VMS apps natively under OSX would have been interesting.

Re:Porting OS X

[2nd Post!]

Why would Classic die, and why would everything have to be recompiled?

Code Morphing.

Transmeta laid off 40% of it's workforce.

IBM has DAISY as an open source project.
HP has Dynamo, which Apple could license.

Re:Porting OS X

[Rasta Prefect]

MIPS - A moribund processor (as far as the desktop market goes). Not enough market share and no real hope of ever getting any (unless the PS2 with Linux kit starts to take over the home pc market space ;) Maybe Apple will buy SGI and whatever patents they have remaining?

Oh dear god. This brings to mind the idea of some kind of Spawn of the iMac and an Onyx, or worse yet a big Origin....

Re:Porting OS X

[binaryDigit]

Think Quicktime7 ported to use the Infinite Reality engine. I can see that Onyx2 with purple translucent skins, but dang those blue led's on the hd's would look kewl.

Re:Porting OS X

[jamesoutlaw]

IA-64 seems more likely than IA-32 to me, but some people have suggested [mosr.com] Apple could move to IBM's Power4 line, which is closely related to PPC. How about ARM Processors? MIPS? Sparc? Alpha? Transmeta? Anybody have any other ideas?
I was wondering the same thing. It seems to me that Apple would stick with a RISC chip similar to the G3/G4 that they currently use. Do you think that they would make the switch to an Intel/AMD processor after lauding the advantages of RISC for all of these years? I know next to nothing about processor design and the advantages/disadvantages of each, except for some general marketing-derived knowledge.

OpenStep was available for many different architectures, Intel, SPARC, PA-RISC and probably a couple of others. I wonder if the remanents of OpenStep and all of the other stuff that makes up the underpinnings of OS X and Aqua can easily be ported to these other architectures?

Re:Porting OS X

[IamTheRealMike]

You will loose Classic support though, and everything will have to be recompiled, but MacOS X provides a mechanism for distributing multiarchitecture binaries.

That's something that I haven't seen raised yet. People say "porting OS X to Intel" in a rather loose fashion, apparently without realising that if Apple released OSX/i686 tomorrow, it would be useless. If I went out and spent a lot of money on it (and it would be a lot of money) I would get the standard Apple apps and .... that's it.

I don't think vendors would recompile their apps. Why do I think this? Past experience. Look at Linux, it runs on almost any architecture under the sun, including PPC. The only reason this is viable at all is because virtually all the software has the source available - if you look at commercial vendors like Real, Yahoo! and so on, they invariably only support Linux/i386. Extracting builds for other architectures is like getting blood from a stone, they just don't want to know .

I'm not sure why this is the case. After all, producing a PPC build is only a matter of going out and buying a Mac, installing Linux on it and recompiling. Nothing to it. Perhaps it's because they think, well if we support i386 and PPC, suddenly users of all sorts of other wierd architectures will start demanding builds, and we can't do all of them. I don't know. But I do know that it would take a long time for the number of MacOS/intel programs to reach the number of MacOS/PPC programs. It'd have the chicken and egg situation that killed BeOS, prevented so many other platforms from even taking off, and Linux had to work around by making all the software open source. Without customers, they wouldn't support Intel builds.

Oh and also, when you mention that there is a mechanism for distributing multiarchitecture binaries, I assume you mean you can put different builds into the same appfolder. That strikes me as an extremely dumb idea. Why exactly would a Mac user want their harddisk cluttered up with unusable Intel binaries and vice-versa? Better to do what is already done, and provide separate distributions for each architecture and you choose which one you want at the website/shop.

Re:Good god please!

[cbiffle]

I concur, but keep in mind that Apple has always (afaik) made the majority of their income from hardware sales. Were they to move to an easily-cloneable PC platform (i.e. the platform that's evolved from the original IBM PC), they'd have to really move a lot of software to make up!
HOWEVER: a lot of people seem to think that that fact alone will prevent them from considering Intel/AMD. I'm not so sure. Granted, all of us would love to be able to run OSX on our existing Athlons, and I doubt that will happen -- but what's to stop Apple from rolling out their own version of the PC architecture? There's a lot of cruft in the ex-IBM PC that could be dropped; with a little work, an x86 machine could be had with all the hardware integrity of a Mac, and just as hard to clone.
You may be asking, Why would we want THAT? Two reasons, imho: 1) it would allow them to lower the price (though I doubt they would), and 2) x86 binaries could be run through something closer to Wine than a full-out emulator! Mmmm, compatibility.

Abandoning Motorola, not necessarily PowerPC (?)

[Etcetera]

Okay, I may be wrong in this (too lazy to check the article which I read yesterday), but I think the original article merely talks about moving away from Motorola.

The implication, of course, is a move to OS X, but I think it's much more likely that Apple will turn to IBM's PPC chips instead. IBM (the other part of the AIM triumverate) has been a supplier of Apple's chips for a while, and they're poised to release the processor's Apple needs well before Motorola (which can't seem to get their act together.

Try http://www.osnews.com/story.php?news_id=1357 or the Thursday, 7/11 update here.

Re:Good god please!

[Phroggy]

Am I the only one who thinks that this should be an *entirely* obvious step that should be taken immediately?

No, plenty of other people who aren't willing to pay for anything agree with you completely. Other people who would be willing to pay up to $129 for the same experience you get from a Mac agree with you too.

Apple is a hardware company. Getting Mac OS X working correctly on commodity hardware would be VERY expensive.

OS X on Intel Processors

[MBCook]

I've said it before and I'll say it again. I would love to see OS X on Intel processors for a variety of reasons. First of all, it's not Microsoft, so that means it's competition. My brother runs it on his Mac and it's great. It looks great, runs great, stable as a rock, and a unix core. I'd put it on my PCs in a heartbeat. Even if it failed, it would make MS push some features taht I think are smart like using 3D hardware to render the desktop to let you do all those cool window transitions and stuff. Also, OS X runs great on Macs, but Intel and AMD are up to nearly 3x times the speed of Macs, so think of what they could do. They could fund Wine for windows compatablility which would be a HUGE boost to open source. Plus software from the Mac could be easily ported, probably just a recompile like most unix software. And all the PC apps that might (and hopefully WOULD) get ported to OS X would easily go back over to the Mac, giving the existing userbase a major reason to want this. Plus the ease of a Mac on a PC would give Apple a reason to lower prices. At this point I'm close to rambeling so I guess I'll end this. I would LOVE to see OS X on PCs. I'd definatly dual boot it, no question. I'd pay $150, $200. And I'd love to be able to access the stuff on my Macs on my PCs without having to pay for some 3rd party program (my Macs are older and can't run OS X, so having an OS that could connect to them would be great). Plus it would probably FINALLY push MS to put in Mac compatibility stuff like Apple's had for years.

Re:OS X on Intel Processors

[Osty]

Even if it failed, it would make MS push some features taht I think are smart like using 3D hardware to render the desktop to let you do all those cool window transitions and stuff.

Microsoft doesn't necessarily need Apple to push them in that direction, since they're already going there. Maybe not with Longhorn, but soon (say, within 5 years, which I guess isn't "soon" in computer time, but it's still relatively soon) you can expect DirectX to be the de facto standard for displaying anything on Windows. No more GDI, nor more GDI+ (which is better than GDI, but is not hardware accelerated by many drivers yet), OpenGL probably just as a wrapper for DX (or at least by default. I'm sure hardware vednors will still write OpenGL ICDs if the demand is there), etc.

Of course, once that happens, expect to hear even more bitching and complaining about having such a heavy GUI on a server-class OS (I guess they could stick with the current GUI for their server-level versions and the DX-based GUI for consumer-grade and workstation versions, but I doubt that would happen for consistency purposes). Anyway, by that time, a DirectX-based interface won't be any heavier (relatively) on future hardware than the current GUI is on current hardware.

Re:OS X on Intel Processors

[man_ls]

There's nothing wrong with having a server with a powerful, easy to use GUI.

When you're not doing anything, the GUI isn't going to suck cycles anyway. And if you have a server in such a bad situation that you have to get up from your terminal and actually walk to its rack, *many* racks have a KVM built right in.

Displaying something on the screen would ease troubleshooting, increase response to downtime, and make it more enjoyable for server technicians.

Just because Windows has a GUI doesn't mean it doesn't also have a command line that is easily accessable and readily useable, either.

Re:Good god please!

[SirSlud]

I somewhat agree.

More importantly, Mac engineers are still sane enough to remember that they became engineers in order to give Joe Q Public what Joe Q Public would want for other people had he spent a lot of time and money studying usability. Ie, expertise for a reason. MS engineers seem to be the whipping boys of the MS marketing and customer feedback teams .. they still do engineering, but they dont seem to have any concept that if Joe Q Public were designing cars (remember the car that Homer built?), we'd be better off using bicycles?

Look, I know I'll get a -1, Elitist for this, but in this day and age of specialization and depth of research, dont we think its time the public started making concessions for what they want and start listening to the folks that are genuinely interested in making really great usable software (even if its quite different than what you're used to), instead of assuming that even tho they need my help to find an ISP and to set up the internet they can't neccessarily ask for the right things that will make their tasks easier to accomplish?

Flame away, but just remember that desires are so cheap, you rarely spend any time or effort making sure they are good for you in the long run ...

Re:OSX on x86

[dhovis]

They would likely still have their own, special BIOS and architecture

I suspect Apple would require a different BIOS than what is used, per se, but right now Apple uses Open Firmware, which is an open standard for booting. You can wipe MacOS off of the so called "New World" Macs (most Macs since the original iMac) and install Linux if you like.

Re:Good god please!

[BoBG]

Lets be clear about this too:

x86 CPUs != IBM PC compatible (necessarily)
Opteron* x86-64 != IBM PC compatible (necessarily)

(or whatever they are calling this chip)

Just because the CPU is an x86 and therefor cheap, does not mean that the system is an IBM PC. Apple could still be a 'hardware' company and move away from the PowerPC, which, while a fine processor with excellent performance can't meet intel or amd in a price/performance showdown.

Apple should make the smart market decision and pick up some of the 'low price' users.

Re:Talk to you all later.

[binaryDigit]

People often attributed Kurt Cobain's famous quote to him: "It is better to burn out than to fade away."

Ole Kurt was not the source of the quote though, that would be Neil Young from around '76-'77 time frame. He just, uh, borrowed it since it, uh, fit the moment, I guess.

Re:DjVu libre link

[jkramar]

For GNU/Linux, the link in the posting leads straight to a free software project DjVuLibre which handles the DjVu format (encoder, decoder, browser plugin, and standalone viewer). There doesn't seem to be a gimp plugin or any kind of editor. For Windows and Mac, LizardTech provides a browser plugin. I haven't seen it before either, but it seems to have potential.

Paying to see ads is just wrong

[MrResistor]

Is it just me?

The advertisers should be the ones paying. Let them post their add and bill them for the bandwidth (plus markup to cover overhead, obviously). It's got to be cheaper than getting TV ad spots, and the advertiser gets direct feedback on how many people actually watched it, as opposed to a guess based on "ratings systems".

Adcritic was one of my favorite sites back in the day, but there's no way I'm going to pay to watch ads.

Re:Paying to see ads is just wrong

[Chanc_Gorkon]

Um...u pay to see ads every day if you are a cable subscriber and sometimes in more ways then one (pop up ads thru road runner any one??).

Re:Paying to see ads is just wrong

[MrResistor]

No, I pay to see TV shows. The ads are something I must endure in the process. If there were a "premium" channel that just showed ads, and I subscribed to it, then I would be paying to see ads.

Re:Paying to see ads is just wrong

[startled]

No kidding. There's been so much doom and gloom about the ineffectiveness of internet ads and how no one pays attention to them. Here they have an audience much better than TV-- they're interested, they WANT to watch your ad, they aren't getting up to go to the bathroom.

If the advertisers get any goddammed sense, they'll foot the bill for adcritic to host their ads. At that point, I will go out and watch them. Until then, I won't pay for adcritic, and I'll skip all their TV ads with my Tivo.

Re:A jpeg replacement - IE does transparency

[tfinniga]

Their support is un-unified at best, but it does work. It's probably not the default just because of performance, or it was a late feature or something. But it does work, and it's pretty cool.

See:

The wrapup of how all browsers do at supporting PNG
A demo of transparent pngs that works in ie for windows

Re:OSX on x86

[Tablizer]

Of course, _nothing_ says that any apple x86 computers would in any way be compatible with standard PC offerings. They would likely still have their own, special BIOS and architecture...

You mean my dream of cool 3D gel-like BSOD screens are out of the picture? Darn!

Ooops, I forgot about the slashdot BSOD joke morritorium.

OS X on Intel? Doubt it.

[kitzilla]

Jobs is needling Motorola. Unless Apple's business model is changing from that of a hardware to a software company, they'll stay with PPC.

When folks see my iBook, they think of it as a "Mac." A Mac is different from a PC (in marketing terms). This difference is why Apple can turn a profit these days when Gateway is posting losses.

If you put OS X on Intel, every beige box will be a "Mac." The name will lose all meaning, and Apple will have surrendered its hardware's marketing position.

It might be that Apple has, indeed, decided the hardware market is too saturated to assure the company's long-term profitability. This is the only reason it would make sense to port OS X to Intel.

I do not agree that the market is tapped out for Apple. If I were Jobs, I would constantly press hardware requirements through technological innovation on the OS and clever new add-on devices. This will keep their existing customer base on an upgrade track. A hot OS and new features, properly marketed, will also serve to attract new users. Their entry point is a hardware purchase.

Given Apple's commitment to their new retail stores, I'd think they still believe they're a hardware company. No Intel for now. Just options. :-)

Re:OS X on Intel? Doubt it.

[captaineo]

If Apple switched to x86 chips I'm sure they would add some proprietary BIOS or something to prevent the OS from running on non-Apple PCs.

To its credit, Apple already has one successful CPU architecture shift behind them (680x0->PPC). So they have experience dealing with the compatibility issues. (though said experience included writing a 680x0 emulator for PPC; it'd be a big project to do a fast PPC emulator for x86...)

Re:OS X on Intel? Doubt it.

[evilviper]

It might be that Apple has, indeed, decided the hardware market is too saturated

I sure hope not. I'm not an Apple fan, but I'd hate to see them go bankrupt.

I've heard a lot of people saying how Mac OS would push Windows off the desktop if it ran on x86... Before anyone even thinks about that, remeber that just as many people were recently saying that Windows users would switch to Linux en-masse as soon as a free Office suite was available. It just goes to show that people don't know what they want, and that goes double for Mac OS on x86.

You know what I think? Leave the future of Apple to Jobs. He's done great so far, I don't think that anything any of us has to say really matters.

So many alternatives

[dghcasp]

Here is an open alternative to the JPEG file format.

Wasn't PNG supposed to be the "Open Alternative?"

Don't hold your breath

[BeBoxer]

Even if Apple switches to x86 procs, their hardware will still be proprietary. Why? Because they would go broke in a heartbeat if they tried.

1) Apple is largely a hardware company, and one with fat margins to boot. If they tried selling PC clones with similar margins Dell would take them to the cleaners. Hell, Dell would probably take them to the cleaners even if they charged slim margins.

2) That stable as a rock feature your brother enjoys? The almost seamless integration of most hardware into the OS? Those are features of the tight control that Apple can exercise over their hardware. If you think you would get these same features running on generic PC hardware you are sorely mistaken. Most vendors don't bother writing OSX drivers now, despite the fact that all PCI, AGP, and USB devices will plug right into a Mac. What makes you think they'll bother writing OSX-x86 drivers? Or were you just going to use the high quality BSD 3d acceleration video drivers? The world of PC hardware is a tar pit of cheap hardware, poorly documented interfaces, and Windows-only drivers. Hardware detection and configuration has never really been one of Unix's strong points. Why do you think OSX would be much better?

And don't let the fact that PIV's have almost 3x the clock of a Mac fool you into thinking it has 3x the performance. The PIV is first and foremost a high speed oscillator circuit. It is designed to have a high clock speed because most people are stupid and think it means fast. Meanwhile, Intel's highest performing chip at FP (the new Itanium's) is clocked slower than a Mac. So is it slower than a Mac? (Not that I'm arguing a PM is faster than a PIV, I just don't think it's a factor of three slower.)

Re:Don't hold your breath

[SirSlud]

>Apple is largely a hardware company, and one with fat margins to boot.

See, I see it as a wierd (and unfortunate situation). I really dont believe Apple couldn't handle a wider hardware market. Drivers are up to the companies that make the hardware, so really, its the other hardware companies that would have to write the drivers. OSX, being built on *BSD, is still going to be inherently more stable than Windows, even if you factor out the hardware situation. My opinion, of course, and potentially groundless.

But I see Apple as a software company that has no choice but to pay for the software via the hardware .. OSX is the bomb (I'm surprised to see all the "I dont run macs, but damn I want OSX" posts, especially given the Aqua factor) .. and the fact that Apple so vohemently protects their look and feel suggests that they put much value in the fact that people know, everytime, that they are running an Apple OS. (Plus, they plug it in movies alot more than MS although maybe thats cause MS refuses to let Windows go into movies? .. I think Apple wants people to see their software, not hardware.)

So anyhow, Apple knows that if they let their hardware monopoly go, they risk not being able to deliver their software at a competative price-functionality/performance ratio. Wasn't that one of the big reasons behind them clamming up on the clone front?

Can anyone give me some hard facts about whether or not its true that they rely on their hardware markups to drive the development of their kick-ass software?

And yes, if anyone wants to diss me for calling Apple a good software company, the end of the Apple OS9 line was crap, but then again, nobody bought Windows ME for exactly that same reason - no amount of money or brains could keep the codebase under control.

Thoughts?

Re:Don't hold your breath

[jamesoutlaw]

Can anyone give me some hard facts about whether or not its true that they rely on their hardware markups to drive the development of their kick-ass software?

I can't give you any hard facts, but II can give you a bit of empirical evidence.

Apple give away nearly all of it's software. They sell a few of their more "professional" titles: AppleWorks ($79), DVD Studio Pro (somewhere around $1,000), iMovie ($50), Final Cut Pro (~$1,000), WebObjects (~$700 plus some cllient fees maybe), OS X (~$130), OS X Server ($1,000) and a couple more. It seems tha that the price of these titiles are comparable to the prices of similar products and that the price would reflect the development costs of the software plus all of the overhead associated with it. Given the quantity of "free" software they put out and the quality (iTunes, QuickTime, iMovie, iCal, iSync, etc) it seems to me that they would subsidise the development of the free stuff from the sales of their hardware. I have no idea how many developers they actually have working on these titles, nor how many people would be required to put together some of these things.
Anyway, that's a long blurb about nothing and it may not make any sense, given the fact that i am pretty tired right now, lol...... any other ideas?

Re:Don't hold your breath

[Sorklin]

I think the alure of apple is to a large part the hardware. OSX is great, but what makes it great is that it works on the hardware solidly, because apple is in control of that.

If they released it to a general x86 environment, that edge is gone and they have to start doing the same crappy thing that MS does with trying to support every damn thing old and new made for x86 platform.

And they can't compete like that.

If they do decide to come out with x86 (maybe to gain some speed increases and take advantage of things like WINE), then they most likely will control the hardware again, making sure that OSX will only run on something with the correct bios. The overall cost of the system may come down, but would still not be competitive to other PCs.

Apple's current systems with OSX just work. I haven't had to fuss at all. Coming from tech background on the windows side, it took a while for me to realize that everything was already setup correctly. Was a bit frustrating at first, but eventually I learned to relax.

Heres a mac feature that I love. If you have a shortcut (or a data file) on your system and you move the application, the mac knows. No broken links or shortcuts. No missing data. Its great.

Gifs???

[robotbrain]

The thing that is blowing my mind right now is how there are GIFs all over the djvu website!??

Re:Yeah, but..

[Lord Bitman]

*opens IE*
*loads his forum, which uses only PNG images*
wtf are you talking about?
IE seems to support .png, please explain what you mean by "IE does not support .png"

Send in the Clones

[mumkin]

Apple had its flirtation with allowing its OS to run on other people's hardware. They killed it off. I'm still using an old PowerComputing clone that I bought back in '95, running OS X on it just fine thank you. Sure, it's been through a lot of upgrades, but until I bought an iBook last month Apple hadn't seen a dime from me for hardware since I bought my old Centris 610 a decade ago.

They killed off their licensing arrangement with the clone makers because Apple makes its money from hardware. It's very hard to imagine that they could sell Wintel users enough copies of OS X to make up for the lost hardware sales they'd get from "switchers" who no longer had to buy an entire new machine. Would it rock if I could run OS X on a tricked out custom-built PC at half the price of an apple box? Sure! Would Apple profit from my doing so? I don't think so.

I suspect that The Steve was just suggesting Apple might switch to IBM's Power 4 as the next gen architecture, not that they'll start dropping Athlons into iMacs.

Re:OS X

[bigfatlamer]

It's already been covered many a time, but DAMN would I love to see OS X on Intel hardware.

Alright, I've been hearing this refrain since OS X Beta first came out and I finally have to ask. What exactly would be so cool about having OS X on x86 hardware? Is there something about the x86 architecture that would make the BSD core run much better than it does on PPC hardware? Or is it just the idea of finally having a usable, decent, attractive OS to load on all that cheap, commodity hardware?

Does anyone actually think that an x86 port of OS X would run faster on a 2.5GHz P4 than it does on a 1GHz PPC? No fucking way. They'd cripple it even if they could get it to run as fast as on PPC hardware, just to give you a taste of the good life...the first one's always free.

E

Re:OSX on x86

[doorbot.com]

right now Apple uses Open Firmware, which is an open standard for booting

Ask a Sun engineer if you want a more accurate response, but as I recall, Macs (including my own) using Open Firmware, which is a stripped down version of the true "open firmware" known as OpenBoot. This is, of course, used by recent Sun machines and is very nice. I wish my Mac had the same OBP as my Sun.

For example, I managed to damage my video card on my Ultra 30. Yes, it was stupid and I regret it. However, I didn't know exactly what was wrong with it at the time. I connected a serial cable to the Sun, fired it up without the keyboard and was able to get into the OpenBoot prompt. From there, I ran the built in diagnostics which basically told me what an idiot I was and how much damage I'd done to the card (doh!).

Now, granted, I have yet to try this with my Mac but I'm 95% sure it won't work.

Time to ditch image files altogether

[guttentag]

or at least reduce our dependency... If you have an HTML page that calls 30 little images, each client that views the page has to make 31 connections to your server. Wouldn't it be better to reduce it to one or two? It would probably speed up slow connections significantly.

It is possible to render images using intricate table coding in which each cell represents a pixel (use colspans and rowspans as necessary to optimize the table).

See my example here. It does use one tiny, two-color gif for the page background, but most of what appears to be images are actually table cells with bgcolors. Unfortunately, it doesn't quite properly in Mozilla, which absolutely refuses to render 1x1 table cells.

In reality, this isn't a total solution, but if image format lawsuits succeed this is what we'll end up doing to render graphics on the Web.

Re:Time to ditch image files altogether

[Monkeyman334]

On my web design site we don't let designers upload designs with images because we want users to be able to use designs without an image editor. People that couldn't break from the images started some pretty interesting alternatives.

Face (IE Only)

Blueish (Ascii pron, all browsers)

Some Impressive CSS Design

2176lines (in emacs, crashes NS4)

C64 (all browsers)

And the list goes on, here's an ascii solution to flash while I'm at it. The web would be just fine without images and flash.

Re:Time to ditch image files altogether

[afidel]

If you have an HTML page that calls 30 little images, each client that views the page has to make 31 connections to your server.

Noy if you are running any modern browser. HTTP 1.1 includes piplining which reuses tcp connections for multiple elements in an html page so you should have maybe 4 connections at the most, though often 1 will do. HTTP 1.1 is the default in all modern version of both IE and netscape browsers including all builds of mozilla since .9.something.

Re:Time to ditch image files altogether

[isorox]

but most of what appears to be images are actually table cells with bgcolors.

Sweet, how did you do the yahoo geocities image just underneath Sorry, this site is temporarily unavailable!?

Re:Time to ditch image files altogether

[guttentag]

Very cool... I have a collection of various "image-free" designs that use table cells and CSS to render pixels.

Have you seen The 5k? My all-time favorite entries are:

1. hungry little frog
2. The Days (IE-only, I think)

Re:Time to ditch image files altogether

[guttentag]

In the source of the mirror file, you can delete everything after the tag. The only reason I have the xmp tag in there is because Yahoo automatically appends code to every file on geocities and I wanted to declaw their Javascript and cookies.

Re:Time to ditch image files altogether

[guttentag]

I just looked at the statistics for that page... yahoo says it only got 14 page views... if 14 page views are all it takes to slashdot someone's low-bandwidth geocities site, that's REALLY sad...

Re:Time to ditch image files altogether

[cygnus]

The web would be just fine without images and flash.

yeah, if you don't mind that sometimes it only works in IE or crashes your browser. geez.

Re:Time to ditch image files altogether

[Monkeyman334]

The problem is he uses an empty 1x1 table cell to render an image using table. The face would render in mozilla if a.) mozilla rendered empty 1x1 cells, a problem mentioned in the parent. b.) He was allowed to use an image, and decided to put in a 1x1 clear pixel. The crash one is just a problem with NS4 not being able to handle 2176 lines and a whole bunch of tables.

OS X on Intel? Maybe!

[Slur]

IT is not out of the question for Apple to make their own computers utilizing Intel processors. They would only support the OEM hardware, avoiding the pitfalls of Windows, which must support zillions of configurations.

The most challenging hurdle I can see is dealing with big versus little-endian issues.

Re:OS X on Intel? Maybe!

[sl3xd]

Actually, endian-ness isn't even an issue either. Both PowerPC and Penium processors have an endian flag-- this allows the processor(s) to use either byte order with no performance drop. Of course, the two have opposite 'native' modes, but it honestly doesn't matter in terms of speed.

Re:OS X on Intel? Maybe!

[ivan256]

First of all, x86 processors do not have a big endian mode. They are strictly little endian. Secondly, the little endian mode on PowerPC has never worked 100% correctly.

Either way, it's a moot point. Good code works correctly on either platform. Even OS code should be written endian clean, since no matter what the endianness of your processor, the PCI bus is little endian, and ethernet is big endian. The MacOS X kernel already runs on x86, so I assume that they've been careful.

Apple should consider buying the Alpha technology

[Tjp($)pjT]

It is scheduled to be retired by HP/Compaq/Digital glom after the next two generations are out. The architecture has lots of headroom and Apple could own it outright. It has been 64 bit from the start and is very flexible. And the Apple^H^H^H^H^HAlpha architecture equivalent of AltiVec (MVI) is quite suave.

I use PNGs exclusively

[AKAImBatman]

All images that I create are PNGs. Check out http://www.datadino.com and see for yourself.

Re:I use PNGs exclusively

[AKAImBatman]

Actually, I did check it when DataDino was created. The only problems have to do with the nowrap tag (non-standard, but supported everywhere) and the lack of TBODY (which is only a sin in the strictist of conformance). Did you have a problem rendering the site? If so, what browser?

Re:OSX on x86

[jimbolaya]

But how would porting OS X to Itanium 2 in the megahertz race? We've all heard Apple preaching about the "MHz Myth," and they'd have to keep on preaching if the went with Itanium, which is, like the G4, settled around the 1GHz mark. Now, I don't mean to suggest that the Itanium 2 isn't faster than the G4, or a 2.5 bajillahertz Pentium 4, but that's just the point: Fighting the MHz Myth with Joe Consumer. I think if Apple did look past Motorola, it would be to get on the GHz train, to cash in on the market buzz that comes with these high revving chips.

Besides, what would they do for their portable line? Itanium is meant for high-end servers, not lower powered laptops.

best commercial ever

[jcsehak]

VW's "Milky Way" (last one on the page). Like the best silent films, they manage to tell a wonderful story with no words, except this is all in the space of a minute. The music chosen gives further meaning to the video, and vice versa. The fact that it's an advertisement is almost an afterthought. Which, ironically, makes you want to buy one even more. I consider it to be one of the finest short films ever made. If more commercials were like this, I'd pay for that adcritic subscription. Especially if they offered a way to get broadcast-quality copies of these commercials.

Re:best commercial ever

[MKalus]

Do you know who created the music?

Re:best commercial ever

[khendron]

I disagree. VW's "Milky Way" is the second-best commercial ever. The best is VW's "Synchronicity" commercial which, unfortunately, does not appear to be available on the VW site.

If anybody's knows where I can get a copy, I'd be grateful.

Re:best commercial ever

[x136]

I agree. I normally despise commercials, but I can watch that one over and over. It's the only commercial that I can think of that actually makes me feel like I want the product. Even the most entertaining of commercials get really old and annoying after you've seen them 10 times a day, but not this one. Then there are the ones that are annoying or stupid from the beginning, and make you never want to purchase a product from that company. I wouldn't hate commercials as much if more commercials were as pleasant and effective as this one. Bravo to whoever Volkswagen's ad agency is.

Re:best commercial ever

[jeffehobbs]

Yeah. That's Nick Drake. Great stuff -- so good that (as a card carrying music snob) I'm embarrassed to admit I first heard of his work from the VW commercial. He's a folky 1970's singer-songwriter who lived a pained and short existence, and finally committed suicide via pills I believe.

Personally, I'd check out his "Pink Moon" album first, and maybe work backwards frm that; and there's a few compilations out now that are truly excellent as well. And there's a short documentary on his life called "A Skin Too Few" making the festival rounds which is supposed to be very good as well.

~jeff

Re:best commercial ever

[dasmegabyte]

It was the Speed Racer GTI commercial that roped me into purchasing my Passat (I am much too large a personality for such a tiny car as the GTI). In fact, there's a Trixie action figure half melted in my warm air duct...

Re:best commercial ever

[Artifex]

I don't think most of us in the United States will ever see the best commercials there are - in every compilation of international commercials I have seen, the Europeans trump us easily, especially when it comes to "public service announcement" type ads. Since these compilations get shown every year at film festivals, it would make perfect sense for AdCritic to offer them as a pay feature.

I wouldn't say it's the best one, but one of the ones I remember best from these annual shows was one that only ever played at one theater in England, for a reason that will be obvious: it was a PSA, and showed a homeless (drug addicted? alcoholic?) person, the quiet kind most ignore unless they are causing trouble, only in this ad you hear not only about the man, but from the man himself. The PSA turned out to be for a local charity that assisted people like him, and at the end of the part where he talks about how they helped him, he says they even gave him a job. The commercial ends with the charity saying that besides just helping him, they hired him to collect donations... and he's the guy everyone outside that theater in England just all passed by and probably ignored on their way inside, as he sat on the ground outside the theater. If I remember correctly, it closes with a shot of him there, waiting.

I think most people in the theater watching the compilation with me were crying at the end of that one, and we were a few thousand miles away. It was... very effective in getting people to care about someone, flipping it back around to point out that they may think they care, but they have ignored obvious opportunities to help, and then pretty much pushing them directly back towards that opportunity.

I'm sure I've really mangled the content, and you can't possibly really get the full effect from what I said - which is precisely why AdCritic should be showing foreign ads, as well. There's a lot of powerful stuff out there, and we should be able to access it online.

Re:best commercial ever

[MKalus]

Thanks.

The name rings a bell and I heard some of his work but I didn't recognize him in the ad and nobody else could tell me.

Thanks, I'll see that I can hunt down some of his Album.

Michael

Re:best commercial ever

[Artifex]

This almost sounds like Barney's little movie in the Simpsons episode where they are doing a local film festival. It was something like "Don't cry for me, I'm already dead". I wish I could find that video.

Certainly, if you mean "almost" in the sense of nothing much like, actually, except it made me sniffle, too =) When they made that episode, I think they embraced more than parodied "art" films with it, and it worked really well.

Motorola and PowerPC processors

[tristan-b]

The funny thing is, Apple hasn't been using PowerPC processors in most macs for some time now. That's right. If you open up your brand new G4 tower, that piece of silicon won't have the PowerPC label. In fact, no processor made by Motorola is made under the PowerPC name. Due to a bit of a hiccup with IBM over licensing the name, they're all merely "PowerPC Compliant" processors. Seems we won't see a PowerPC G5 coming from them after all.

An Intel Mac: Different Beast

[Spencerian]

Don't presume that moving to Intel hardware will create a Mac with the highly-modifiable box you take for granted on PCs.

Apple survives today because their boxes are designed to make a user's life easier. That means, despite a change to the processor, it is very likely that Apple would still have a custom motherboard available ONLY from Apple, still use Open Firmware rather than a PC BIOS, (this is done on Sun as well) and still not be subject to the resource-hungry design of the aging PC design.

Intel may assist Apple in a mobo design, but Apple will not release it for general consumption. If they want to continue to survive as a business, it would be suicide to do so. Apple is a hardware company. They have to keep some things closed to keep a competitive edge. The hardware would be generally closed-source, along with the upper layers of Mac OS X (Darwin, the core of OS X, is open source and works right now on x86 as well as PPC.).

A more serious matter would be the Pentium's lack of Altivec--the vector processing unit and the true power in the PowerPC chip that lets it keep up with Pentiums doing the same calculations in most instances, despite PPC chips having half the clock speed.

Not insurmountable things, however. I tire of the PowerPC production issues at Motorola. I would rather get IBM to make the chips--they should know how, since the PowerPC chip uses the same tech as in the POWER mainframe chips.

Re:An Intel Mac: Different Beast

[Spencerian]

Yes, I'm fully aware of the PowerPCs origins. My comment doesn't imply anything otherwise, it just wasn't specific. IBM didn't solely create the PowerPC spec, however--this was created through Apple and Motorola as well.

I don't have enough knowledge of Intel's vector processing abilities, so I'm not setting up a debate on this, but need to ask a few questions.

Is this SSE2 built in standard Pentiums, or only in Itaniums or other processors from Intel?

If this vector processing is available, why does a PowerPC G4 Mac running Photoshop tend to match or exceed the processing time in accelerated functions on images in Photoshop on Intel hardware (these are Pentiums used in the various informal tests that Apple and others have done)? Is Photoshop not optimized to take advantage of SSE2 (or the particular vector code available) so that, given its larger clock cycles and system bus than the current G4s, the Intel chip should whip around a G4 handling the same tasks as if it were standing still? This, of course, presumes that Apple and others don't "cook the books" in their demonstrations.

If Intel's vector code is as relatively convenient and as powerful as you describe, then this is a good thing should Apple consider a move to Intel processors--without this acceleration, there's not much of an advantage for their hardware.

Re:An Intel Mac: Different Beast

[connorbd]

A more likely rumor I've heard is Apple switching to a POWER4 variant rather than stay yoked to the broken Motorola core.

I think the only thing holding Apple up in the MHz race is their commitment to the G4; they're probably wishing they'd gone with the Philips TriMedia at this point. As I understand the situation, IBM can produce the parts Apple needs to keep up in MHz, but isn't allowed to because of contractual issues with Motorola regarding AltiVec. We could have 2GHz G3s at this point (NB the PPC 750/750CX is an IBM part, not Motorola), if Apple wanted them (and I'm sure they'd use them if they could get away with it, but marketing likes the illusion of Progress).

As for proprietary hardware: Apple couldn't get away with doing that. As pointed out, the core of the system is open source to begin with, and the modern Mac is a lot less proprietary than it looks anyway. They have some interest in doing this, true, but in the current climate I don't think they could justify spending the money on anything more elaborate than Yet Another PC Chipset.

(That said, I'd love to see an Apple BIOS in action...) /Brian

Why you don't need to switch

[ajs]

You don't need to use something to replace JPEG... probably. Realistically, the people they're going to go after for JPEG use are the manufacturers of products that use it (digital cameras, image editors, etc) and make a lot of money doing so. This is why Sony was an early target.

If all you do is take an image out of your camera and put it up on your site, you're not infringing their patent anyway (IANAL, so go ask yours), the camera maker is.

I suggest that everyone take a deep breath and then those few of you involved in deciding what image formats are used by open source software should get together and start working on a long-term solution. JPEG is very nice, and if it's still the best technology in town in 17 years (or the only one that's unencumbered), we'll go back to using it.

Re:Talk to you all later.

[slickwillie]

"It's better to burn out than to fade away." -- Kurt Cobain

That's kind of a no-brainer, isn't it?

Star Trek

[bcrowell]

I believe Apple already ported Classic MacOS to Intel. IIRC, the code name was "Star Trek." They never released it. I think this was during their aimless period when they were thrashing around with Copeland and clone licensing. The fact that they never released it makes a heck of a lot of sense, because their business model is based on proprietary hardware. Proprietary hardware means (1) they can make the software Just Work(tm), and (2) they don't have to compete in the commodity PC market.

Apple's hardware strategy made sense back in the days when the extra money you paid for a Mac was really worth it just for all the extra hardware stuff that would Just Work(tm) -- until recently, 24-bit color, built-in sound input and output, etc. etc. were not common in the PC world, and were not cheap either. But within the last year or so, it's gotten so you can get comparable features in a $300 Walmart or Fry's PC.

Re:Star Trek

[DavidRavenMoon]

I believe Apple already ported Classic MacOS to Intel. IIRC, the code name was "Star Trek."

Yes, this is correct. They did use parts of StarTrek for Quicktime for Windows.

Also remember that OS X is based on NeXTSTEP, and NeXT had an x86 version of NeXTSTEP (OPENSTEP), AND Apple also had Rhapsody for x86 (Yellow Box), which is more or less the early version of Mac OSX Server.

There was talk a while back about meetings Jobs had with AMD... the plot thickens.

When asked if Apple would seek an alternative to Motorola's PowerPC, namely chips from Intel, Apple CEO Steve Jobs said the company first had to finish the transition to Mac OS X: "Then we'll have options, and we like to have options,"

Never Saw NeXTStep Intel?

[snookerdoodle]

If you think Jobs is above running on Intel hardware, think again. Intel hardware does not imply beige boxes, anyway...

'NeXT Thing you know, we'll have FAT binaries! ;-)

Mark
'Still has his Framed '88 NeXT Poster On The Wall

Re:OSX on x86

[norwoodites]

The serial part will work on macs with serial ports, in fact that was the only way to access the OF for the longest time on a mac because, the OF drivers for the video card sucked.

OS X on Intel

[dr00g911]

At several times during past keynotes, Jobs has mentioned the possibilites of using new chips from both Moto and Intel.

I've personally heard him mention the possibility of transitioning quite a few times over the past few years -- usually mentioned as an afterthought.

Because of Jobs' pacing and delivery (and his famed RDF), the media hasn't picked up on it very much -- and whatever stir is caused dies down very quickly, and people forget again by the next time an expo rolls around.

I imagine during the last couple of years of OS 9, this was more of a veiled threat to Moto to try and keep up with clock speeds and the like, much like Apple's (non-veiled) threat against ATI.

Honestly -- with the exception of the CPU -- just about everything in modern Macs is a standard across the industry.

Now with Darwin for X86, it shouldn't be terribly difficult at all to transition to Intel -- and might be a welcome change of pace. I've had a 450mhz G4 tower for 4 years now, and machines have barely doubled in speed.

We've heard rumours of the G5 for, um like, 3 years now? Always 6-8 months out (just like today).

I'm as much of a Mac nazi as they get, but as long as the machine behaves the same and -- God forbid -- prices might drop a bit, I'm all for it.

The Mac is all about the user's experience. And, for the most part, the user could give a shit what's in his box as long as it behaves consistently and reliably.

Re:OS X on Intel

[Anthony]

Don't forget. When Jobs' NeXT jetissoned it's manufacturing plant and limped along as a Software/OS company, it branched out into NeXStep for Intel. I could never afford/justify spending the money on a 486/66 required to run it though.

Re:Apple should consider buying the Alpha technolo

[Rasta Prefect]

It is scheduled to be retired by HP/Compaq/Digital glom after the next two generations are out. The architecture has lots of headroom and Apple could own it outright. It has been 64 bit from the start and is very flexible. And the Apple^H^H^H^H^HAlpha architecture equivalent of AltiVec (MVI) is quite suave.

Too late - I'm pretty sure Intel picked up most of the intellectual property involved, as well as most of the Engineers to work on IA64.

JPEG2000 is royalty free

[yerricde]

JPEG2000 is more encumbered by patents than GIF. It'll never see the light of day.

The latter does not follow from the former because those companies who currently claim patents on part 1 of JPEG2000 have also agreed to license their patents to the general public without royalty, unlike the recognized owner of LZW (GIF's back end) and the apparent owner of RLE-plus-Huffman (JPEG's back end).

Re:OS X

[KoopaTroopa]

What would be so cool about OS X on x86 hardware? The exact same things that make OS X as cool as it is on PPC hardware. OS X is just a wonderful operating system, regardless of the hardware behind it.

The fact is that x86 hardware has a huge installed base. Given the assumption that such a port of OS X didn't require any proprietary motherboards or whatnot, (an assumption that may or may not be reasonable,) having an x86 port would be a wonderful opportunity for more people to run this great OS.

On the subject of speed, my eMac is certainly not top of the Mac line and it performs very well on everything I do on it. I'm extremely pleased. The 10.2 Jaguar release will only improve this. Even if Apple was dastardly enough to artificially cripple an x86 OS X port to run at the same speed, I'm pretty darn happy as-is.

Re:Good god please!

[Maserati]

There are two basic problems with the otherwise brilliant idea of just slapping an x86 chip into what was a G4 case (modulo cooling). The first is the network effect. If OS X86 runs Windows apps just fine, then relatively few developers will be tempted to do use the (far superior) Cocoa/Quartz APIs. The other problem is that if Apple ever ships a machine capable of running Visio out of the box, then Microsoft will go into a hyper-adrenal fight-or-flight state and do their best imitation of a rabid badger on crystal meth.

The first problem marginalizes OS X and turns Apple into Dell or Gateway. They're much too good of a software house for that. The second problem will produce a reaction from MS that even a GOP DoJ can't ignore.

Given that, the concept of telling Motorolla to put up or shut up must be attractive to Steve Jobs. He might well float rumors to keep Motorolla on their toes, or to get them there in the first place.

I'm sure Apple's hardware engineers can handle the task of cooling a P4. The other components won't be affected much more than a whole new mobo would normally require. But you'd lose the vector unit and the the lower power requirements of the G4. I don't think they'll do it, since the 'books have good battery life thanks to the lower power consumption int he chips. And that's a strong selling point.

Re:OS X

[MoneyT]

How could they really keep it proprietary enough to avoid the price wars? Honestly, the only thing that isn't standard PC hardware in the macs now are really the motherboard and the processor. That's the only reason you can't build your own mac (that and you wouldn't get the cool case). If they switched to a consumer availible MB and Processor set, they would have to become price competative. That's not something apple really needs I don't think, but I'm not an economic analyst either.

Talk ot the Town

[LinuxMacWin]

I have moderator access, and I am wondering if I should write a comment (and hence can not moderate), or moderate (and hence can't post). However, noticed something which has not been represented so far. Hopefully this will not go unread as there are already 200 comments posted.

People say mac users are fanatic (and that's why they don't switch). People say Mac is losing market share. People are unhappy with Apple charging $99 for DOT-MAC. And so on and on...

However, given a choice of 4 subjects (JPEG alternative, patent infrigement, ad critic, and Apple), I see most of the comments (more than 60% per my count) are about Apple. Either that reflects the zealotry on the Mac side, or, boy, is Mac the talk of the town !

Unbelievable...

Re:Talk ot the Town

[Sorklin]

I've been noticing this for a while. Take a look at the Apple articles that hit the front page (not the ones that are in the Apple section only). They consistently get a higher number of comments than most of the other articles that I see.

I have a Mac and love it. I'm a switcher (though I switched before those commercials) because I wanted to work on FCP. I love every minute of mac.

I haven't been this excited about computers and computing in general for a long time.

Get the crack, come on board.

Re:OSX on x86

[artemis67]

It would be pretty cool. Connectix would modify their VirtualPC software so that you could run Windows (natively) as a process within OS X, much as VirtualPC for Windows does now; the option to run both OS's without the speed hit of emulation.

The biiiiiiiig downside to going with x86 chips... Classic/OS 9 apps would not run. Bye-bye QuarkXPress, and every other big app that developers are dragging their feet on.

And another problem, the fact that all existing OS X apps would have to be recompiled for x86 binaries. Not a major task for the developer, but is every developer going to jump right on it? What a mess, after just having gone through the OS 9 to OS X transition.

And then there's the whole issue of hardware. No more fuzzy math talk about how megahertz don't count. Apple would have to get damn competitive on hardware, either by offering very competitive pricing with narrow margins (not likely)or by continually trying to add features that PC vendors only talk about (like they did with USB, FireWire, DVD burners, and screwless, hinged cases).

Apple wouldn't dare make the x86 jump for at least another year, minimum, and two years at best. They need to give the Mac community a little breather as we just went through the PC equivalent of open heart surgery.

Intel Apple does not mean Apple cloans

[nedron]

If Apple ever did switch CPUs (I don't see this happening with the investment all their big name software vendors have in AltiVec code), it would be on their own machines. You'd still have to buy an Apple to run OS X. I own a couple of nice G4s and have no complaints about the price or performance.

There is a concrete example that exists to show why it would be a terrible mistake for Apple to move to Intel hardware in general, rather than Apple built Intel hardware --- Be. When Be was concentrating on building hardware (does anyone remember the BeBox) and software to match, they were in good shape. It was only when they dropped their PowerPC hardware and moved to generic Intel that they began to have problems. Distribution channels are locked up by Microsoft licensing (this won't change), they had to deal with everyone's $10 NICs and video cards, etc.

Re:Intel Apple does not mean Apple cloans

[anothy]

When Be was concentrating on building hardware... ...and software to match, they were in good shape.

oh? by what measure? how many people bought BeBoxes? i did, but i've only ever met two other owners, and i work with geeks. sure, Be managed to hang on long enough to reach IPO, when the dot-com bubble was still inflating, but you've got alot of work to do to demonstrate that they were doing well by any means.

Re:Intel Apple does not mean Apple cloans

[nedron]

You'll note that I never said Be was doing well, just that they were in good shape. Had they continued their plan for selling their proprietary hardware/software to the content creators, they would have had a niche captive audience the same way Apple does.

Be produced some very interesting OS features, but what did it buy them in the face of the MSFT juggernaut? Nothing. Had they switched CPU architectures but, and this is the key, still built it themselves and that was the only way to run BeOS, I don't think they would have imploded as rapidly as they did. They couldn't even give BeOS away free to manufacturers for a couple of good reasons. Those manufacturers were/are/will be under significant contractual obligations from MSFT that make it economically impractical to provide a competing OS. Secondly, when your average Joe walks into the store, he's going to want a Windows machine (unfortunate as that is) because that's what his friends have and his friends have offered free (pirated) software. If this weren't the case, why didn't people flock to OS/2 when it was available as an option? It was arguably better than Windows, yet Joe Sixpack had no interest in it.

It all boils down to one thing: when you play in the Intel world you're playing in Microsoft's sandbox and they are more than willing to play the bully. I don't see that changing.

Re:HTTP 1.1 pipelining

[guttentag]

I guess Mac IE doesn't understand HTTP 1.1, because I often find that the browser will get stuck in the middle of loading a page with the status bar indicating that it's trying to download a spacer gif from the same server. It just sits there for 5 to 60 seconds. Clicking STOP and RELOAD is usually faster than allowing it to finish. The issue comes up at least once per browsing session, but never seems to occur in other browsers.

Re:Good god please!

[MaxVlast]

Someone's been saying that since the first Mac came out in 1984. The debate will probably continue for years.

IBM PPC's

[Genady]

While it's fun to speculate about Apple going to X 86, it's more likely that they'll shift to IBM PowerPC Chips, perhaps the single processor POWER4 that they've been working on. The major thing that's tying Apple to the G4 right now is AltiVec, for the acceleration of Quartz Graphics. However 10.2 releases QuartzGL, which off loads some of the quartz rendering to the GPU. If Apple can offload enough of this processor intensive work to the GPU then they have reduced their need for AltiVec (which is a Motorola PPC only technology) and opened the door to going with POWER4. Now considdering that POWER4 right now comes 2 CPU's to a die this could be a big win for Apple in the workstation arena. a Dual processor POWER4 machine (for an actual total of 4 processors) with the right graphics card could give the SGI Fuel a run for it's money, especially if the claims about Apple's fast IDE can be believed. Or if RapidIO/InfiniBand ever materialize.

Re:OSX on x86

[gabeman-o]

First of all, just because Apple might switch manufacturers does not mean they will switch to x86. Such a move would not only be impossible, it would be foolish. If Apple were to have anyone else develope processors for them, it would be IBM, as they are the other part of the PowerPC. Neither Motorola nor IBM seem to have an interest in making consumer level processors, which is why it is so long between updates. Motorola also (I believe) owns AltiVec, so if Apple was going to have someone else manufacture for them, they would need a processor that could perform significantly better with much higher clock rates to compensate for the lack of AltiVec or pay an arm and a leg for processors since the manufacturer would have to pay licensing fees to Motorola.

Re:That would kill GIMP

[ajs]

This is all true, but taking the long view, I'll be happy with the current state of affairs if we're able to get some patent reforms in the next 10 years. Open source will still be around in 20, 30 or 100 years. Will this little two-bit company? Probably not.

Re:Abandoning Motorola, not necessarily PowerPC (?

[jamesoutlaw]

One potential problem with Apple going exclusively with IBM is that IBM does not manufacture G4 processors- just the G3- and all of Apple's current machines, except for the iBook use the G4. I read that they are quietly phasing out the low-end G3 iMac in favor of the eMac.
Motorola owns the patents (or some of them) on Altivec (Apple's "Velocity Engine")... though some people have wondered if IBM could simply license the technology and run with it. I think also, that a couple of years ago IBM and Motorola hhad some philosophical differences of opinion on the future of the PPC and the AIM "allliance" sort of dissolved. Apple sells quite a few computers though, over 800,000 last quarter.... that might be attractive to some of the alternative processors.
Also, there have been rumors floating aroudn about Apple simply buying the rights to the G4 from Motorola... regardless of what's true.... Steve Jobs was correct, they have plenty of options.

Re:Good god please!

[Technician]

Apple is big enough to ask Intel for a custom chip just like Microsoft did for the X-Box. (notice the custom socket on the chip?) If they do not infringe on any of Motorola's patents, it is possible for Intel to manufacture and non-X86 chip for apple that would contain all the power, speed, and effecient manufacturing for low cost parts with high performance. What would you do with a 2.5 Ghz Mac?

Re:OS X

[spitzak]

I would think a problem with the proprietary solution is that people will hack the hardware and support software enough that they can get it to work on commodity hardware, I think.

Re:OS X

[DavidRavenMoon]

By the way... if anyone was at MacWorld Expo, and got a chance to try Jaguar (10.2) on any of the Macs at Apple's "both"... IT JUST SCREAMS! It's much faster than OS 9. I can't wait until August!

New G4s should be out in August also.

The 17" iMacs look great. :)

Re:OS X

[DavidRavenMoon]

that was "booth" not "both" :-O

Tell that to Nike owners

[freeweed]

Or hell, anyone who buys Tommy, or damn near anything that's marked up insanely because of the logo. Americans (and Canadians, and Britons, and the French, and the Japanese, and, and, and...) LOVE not only advertising for companies, but paying for it as well.

It's something that sickens me to no end - when I went to high school, people kept the tags on their baseball caps to show they were the *real thing* - that being Starter brand, iirc.

I'm quite happy buying damn near everything I can with generic/unknown brands, and strangely enough, my quality of life hasn't suffered one bit. In fact, I can afford a hell of a lot more pairs of shoes than my Nike-buying friends. And wow! They last just as long. Now if only we had a generic brand of car (and whoever mentions KIA gets a kick in the face :)

Re:There is an alternative to JPG

[Kevinv]

PNG is lossless. JPG is lossy. JPG can make files smaller than PNG (at least at aggressive compression ratios). Websites want the smallest file they can get to speed browsing (ok, at least responsible websites would want the smaller file)

So yes PNG is a great format, especially for intermediate work in a graphics workflow, but a lossy compression is desirable for an end product to get a smaller file.

Agreed, DJVU is awesome... put in Mozilla please

[ecloud]

DJVU was designed for scanning paper documents, mainly. It doesn't include actual ASCII text like scanned PDFs can, but it is tuned for accurate visual representation of printed pages. The images have a couple of layers - the high-contrast material (such as text) and the background nuances (shading of the paper etc.) So it's possible to view the image with or without the background. And the background, being low res smooth shading, doesn't cost much space. I've been using it for a couple of years for this purpose, because neither JPEG nor PNG are ideally suited for it. Separating the foreground from the background strikes me as a really good idea.

It wasn't completely free at first so that was good news when it was announced a while back.

Of course to be truly useful as a JPEG replacement it needs to be included in browsers. People don't like having to mess with plugins.

Anyway I bet the JPEG patent stuff will blow over soon enough. This company obviously has brass balls and no brains to think they can pull that off. Maybe they will manage to sue a few big guys, bully some little ones, make some money and then get a little more complacent as their time runs out. But I imagine we will still be seeing lots of JPGs 10 years from now (just like GIFs didn't go away).

Re:OS X

[Dreamweaver]

At the risk of being flamed from here to Topeka...

OS X on x86 might be nice in that it's a userfriendly OS other than Windows for people to use, but it's not the grail that you and a lot of the Mac-advocates here seem to expect.

Do you know why OS X is such a nice OS? Because it runs on Apple computers. And I'm not saying that Apple computers are better than x86 computers, I'm just saying that they're made and controlled by Apple. They have all the specs on all the hardware down to as low a level as they care to look. They can optomize the OS to run on the PPC processor, to read from Apple HD's, to use Apple CD drives... And none of them will ever do anything that the software doesn't expect. There won't be any weird missing functionality in the drivers. Every hardware glitch or driver bug that may exist is ready and available in a database for the OS programmers.

Porting OS X to x86 would be the functional equivalent of running console emulators on your PC. MAME and the various old-system (genesis, NES, SNES, TurboGrafix...) emulators work fine because modern computers are So much faster than the original hardware that the fact that none of the hardware is at all like what those games were optomized for doesn't matter. Motorola processors may be out of date compared to AMD/Intel processors, but not nearly so much as, say, a 1988 arcade cabinet to AMD/Intel processors.

OS X would be slow on x86. Since it's based on a Unix core instead of some totally proprietary thing it probably would still be easily usable, but it would Not have the performance that Mac-advocates so praise when poo-poo'ing the 'MhZ Myth'. But even worse than the speed hit; it would be unstable. Mac Products 'Just Work' because Apple OS programmers can take the latest build down to a lab, load it on all the machines there and see ever possible major hardware configuration change. The ones that crash, they check the hardware docs and fix the bug. If MS could do that, or if Linux hackers could do that, I have no doubt that Every OS would 'Just Work'.

OS X on x86 (and all this assumes that Apple ports it to open hardware, not proprietary Mac Hardware that's just based on a processor made in a factory with AMD on the sign) would make Apple a lot of money up front, but would be a horrible, horrible idea with their current business model. They Rely on the fact that their hardware is proprietary, and not because they can price it however they want ('cause it's really not That expensive); it's because they have absolute control over how you set up your PC. People slammed MS for giving favor to certain manufacturers, but Apple is worse a thousand times over.

So in conclusion, as I've ranted long enough, if Apple releases an OS X that runs on my 1GhZ AMD, I'll give it a shot, sure. I mean, really, what do I have to lose? And I'm sure a lot of people will do the exact same thing. But you'll have a whole lot of Mac-advocates (Maccies? Macacates?) left sputtering, "Illegal Operation? But..but... It USED to Just Work!"

I wouldn't trust LizardTech

[Kevinv]

I'm guessing they have patents on DjVu, purchased from AT&T.

I use LizardTech's MrSID product for GIS imagery and they are hardcore about declaring (and enforcing) their patents on the wavelet compressions it uses.

That said, MrSID compressed images are amazing tight with very good resolution.

Re:OSX on x86

[BWJones]

I think the Steve comment was taken a bit out of context. Here is the whole quote: "The roadmap on the PowerPC actually looks pretty good and there are some advantages to it. As an example, the PowerPC has something in it called AltiVec, we call the Velocity Engine -- it's a vector engine -- it dramatically accelerates media, much better than, as an example, the Intel processors or the AMD processors... so we actually eek out a fair amount of performance from these things when all is said and done. And the roadmap looks pretty good. Now, as you point out, once our transition to Mac OS 10 is complete, which I expect will be around the end of this year or sometime early next year and we get the top 20% of our installed base running 10, and I think the next 20 will come very rapidly after that. Then we'll have options, then we'll have options and we like to have options. But right now, between Motorola and IBM, the roadmap looks pretty decent. "

Motorola G5 link

[morcheeba]

Here's the page for the Motorola MPC8540 -- the first implementation of the commonly called G5 chip.

G5 seems to be a term that only Apple uses (maybe IBM, too?); Motorola calls this the first implementation of "the e500 high performance core [which] implements the enhanced PowerPC Book E instruction set architecture". They also call it a PowerQUICC communication processor; I suspect they have no shortage of names.

I've been drooling over this chip for a while - RapidIO bus interface, dual 10/100/gigabit ethernet controllers built in, DDR memory controller with ECC, PCI-X... It's got enough stuff to make a great laptop, but is meant more for embedded applications.

Newsflash

[autopr0n]

Unless something has changed, Itanium is a lot slower then current x86 chips out there.

Re:Newsflash

[sl3xd]

Only on x86 code. It's like a mac user (10 years ago) complaining that their new PowerPC runs the same program slower than their old 680x0.

Anything runs slower under emulation -- whether it's hardware or software providing the emulation. Espescially if the emulation is more of a 'white elephant' that isn't really intended to be used.

The Itanium is x86-compatible. There was never a promise that Itanium would execute x86 fast; the current Itaniums aren't even meant for the consumer market, but for workstation use; workstation code, such as the OS (Win64 & Lin64) and apps. And the apps are usually written to be portable, as the market requires it to run on (PPC, PA-RISC, x86, SPARC, MIPS) computers anyhow. Just re-compile.

And, FWIW, Itanium runs PA-RISC code about as well as the PA-RISC does. There's just more in common between PA-RISC/Itanium than there is x86/Itanium.

What?

[autopr0n]

Are you saying PA-RISC and PowerPC are code compatable? I find that hard to belive. Itanium can run x86, PA-RISC and Epic code directly. Nothing else.

Why would intel and apple design a new, even more complex chip, when they could just a java style JIT for PowerPC code?

Ever heard of SSE?

[autopr0n]

Intel chips do have something similar to AltiVec, although I don't know how comparable they are. I'd be willing to bet a 2ghz P4 could probably womp a 800mhz G4 at vector math anyway, these days, so it's really a non-issue.

Re:Ever heard of SSE?

[nusuth]

SSE isn't even comparable, there are lots of thing you can't do with SSE but can do with Altivec. SSE 2 fixes the most important one of these issues, full precision floating point, yet offers little else over SSE. Altivec is about two times faster clock/clock but the huge clock speed difference more than makes up for it if both programs are vectorized. Since Altivec has some features that can only be implemented with relativily slow fpu on x86s, this is not always possible. Also all new Apple software is targetted to two specific CPUs, both with Altivec units. So their programmers are more eager to use all latest features. OTOH SSE2 has a very low adaption rate in x86 world as most cpus do not implement it.

I don't mean to say P4 is not competetive with G4, actually it is way faster. Even faster are Athlons, just not in SSE2/Altivec optimizable stuff (since they can't do SSE2.) If P4 or Opteron based Apples come along they would also be very fast (like they did with G4s, software will be specifically optimized for apple hardware.) I just wrote this to clear up some confusion over benchmarks.

yes.

[autopr0n]

Does anyone actually think that an x86 port of OS X would run faster on a 2.5GHz P4 than it does on a 1GHz PPC? No fucking way. They'd cripple it even if they could get it to run as fast as on PPC hardware, just to give you a taste of the good life...the first one's always free.

Yes.

Anyway, apple will probably build proprietary hardware around an x86 core, rather then just use commodity PC parts. There's no real reason to use PowerPC these days over x86 with all the advancement in emulation and Just in time compilation these days. Some company (HP I think) Actually created a MIPS emulator that ran MIPS code faster then natively, even on the same chip.

You don't know what your talking about.

[autopr0n]

Look, the fact of the matter is, there is not really that much variation in hardware on the PC side. A lot of the problems you see on a PC come from crappy drivers written by companies that don't have enough money to create good ones.

If you buy the right hardware, from the right companies, you can build a very stable windows system (in fact, with windows 2000 you can get very good stability without super-high quality HW) that never, or rarely ever crashes.

Also, if you have any experience with PC OSs other then windows, you would know its quite easy to build systems that don't fuck up and crash using open source Unixes. And what is OSX based on? BSD (along with a Mach kernel)

It wouldn't be anything like trying to run a video game hard coded to a specific hardware implementation. Almost all of the OS code is going to be cross platform enough for you to get it up and running without changing very many lines of code. Of course, you're probably going to uncover new bugs, but those can be fixed. If apple programmers are writing their code so that it only works well on PPC hardware, they aren't doing their job. Sure they are going to be doing platform specific optimizations, but those optimizations can be redone for x86.

correction.

[autopr0n]

My guess that while it will be running an Intel processor it won't run on 'Commodity' PC hardware. The fact that Apple controls the hardware and the software is what makes a Mac so great.

You mean "The fact that Apple controls the hardware and the software is what makes Apple so much money". You can build a PC that's just as stable if you use certified drivers and quality hardware. If apple dropped their exclusive hardware they would be in direct competition with M$. Not a good place to be. Anyway, If apple does switch to x86, they'll probably keep right on with the same business model of proprietary hardware, just with a different chip.

You can get open commodity PPC hardware just like you can with Intel stuff.

Re:correction.

[connorbd]

Could you cite a source for said hardware? I know it's out there, but I wouldn't know where to find it.

I think Apple, should they get into the x86 hardware market (which I don't think would be a particularly good idea), might consider developing their own PC chipset, open to others to develop for, but meant specifically as a Darwin/X platform. /Brian

Well,

[autopr0n]

There was a long time like it seemed like they were going straight to hell. 700 million dollar losses per quarter, etc. Before jobs came back the company looked to die.

I know this is going to get buried, but....

[TellarHK]

Apple won't make Apple machines commodity items. Period. They might move to an x86 (or more likely a 64 bit architecture) core, but they'll only work with Apple systems. Why? Because that's how Apple makes money. They don't make money on the monopoly tie-ins, they don't make money on the OS. They make money on hardware, just like a Dell or a Gateway with style and a sense of personality.

Sure, Jobs is a megalomaniacal asshole a lot of the time, and yes, Apple's as bad as Microsoft in many ways (worse in a few), but despite it all they make good hardware. Hell, even Microsoft makes good mice.

If there's an architecture change, however, this is going to be an absolute pain in the ass for developers. Everything will either need to be compiled again and resold, or it'll all run slower on an emulation layer. Having just gone through that pain with OS X migration, I suspect Apple would have to kiss -everyone's- ass to get them to bother doing it again within the next two years.

Apple's in a very tenuous spot right now, and I'm not sure if they're aware. They're just starting to get some developer approval for X, the OS is just barely coming into maturity, and they've decided to charge a rather high (for Apple users) price for the OS loaded with new features when the current one isn't even working acceptably fast compared to the OS it replaced. They're alienating old-time users, and some recent pre-"Switch" switchers like myself are starting to think the whole give-Jobs-a-chance thing was a real waste of effort because the company just doesn't treat users well.

Dumping "rumor" reporter admissions to MWNY, killing iTools and going pay, charging $129 for OS X.2... It just hasn't been a good month to be an Apple owner. I just hope Jobs realizes this and does something to show us his company gives a shit about the actual buyers.

Though for all Apple's flaws, I -do- like my iBook. It works. Just... a bit slowly.

Unless they pull a RAMBUS...

[autopr0n]

Which could happen, and hold up the JPG2000 standard in court for years.

Hrm.

[autopr0n]

Before it was AMD and Intel, it was Intel and motorolla. If you think about it, AMD's business model was kind of suicidal. "We are going to take on the market leader head on with a compatible product." Most people would try to 'go around' Intel, in the way transmeta did. But AMD stuck it out and came out as a variable alternative. In theory it could happen again.

Also, Apple could use AMDs x86-64 rather then IA-64, which would still give us choice in the market.

I also appreciate that, thanks to the G4, Intel can no longer claim clock speeds are the only meaningful chip performance measurement.

I remember when PPC speeds were higher then Intel speeds, and all the Mac zealots were claming their chips superior based on that. I haven't seen any benchmarks lately either way, but I have a hard time believing that Apple is actually faster. You notice they aren't claming a speed upgrade in their 'switch' ads? Its because they aren't faster.

Re:Hrm.

[Lars T.]

If you think about it, AMD's business model was kind of suicidal.

Was?

I remember when PPC speeds were higher then Intel speeds, and all the Mac zealots were claming their chips superior based on that.

All Mac zealots were saying they were faster because they were faster. All the Introlls said that speed doesn't matter. At least not as long as the Mac is faster. Now they claim PCs are faster (because MHz is everything), and we are supposed to switch to PCs immediatly.

And speed is not an issue in the Switch adds, because for most things (raw processor) speed simply doesn't matter when it is beyond a certain level. Ease of use does.

Nope.

[autopr0n]

Also, it's only the iMac that has an 800 MHz chip. The PowerMac G4 has 1 GHz chips. I'm sure there will be speed steppings down the road if Apple feels threatened by Intel and asks for them.

not according to the apple store. The cheapest mac there with a 1ghz is a dual and costs $3k

The BSOD is not a part of the BIOS.

[autopr0n]

The BSOD is the windows equivelent of a kernel panic.

Whats wrong with KIA?

[autopr0n]

Korean cars are like korean RAM. In fact, Hyundai makes both

MacOS on x86... yawn

[Bud]

The same old zombie discussion "Apple is going to switch to x86" pops up again. Get over it folks. I know that half of you are running Aqua themes on your KDE/Gnome desktops and need to replace your shorted out keyboards every two weeks because of excessive drooling over Apple hardware and software, but hey, really, Mac OS X is NOT going to run on commodity Intel hardware within our lifetime. That would be suicide for Apple. OK?

Besides, it seems Steve was misquoted. Sorry.

--Bud

Re:Time to ditch image files: I have done it!

[guttentag]

I wrote a shell script about a year ago that converts .XPM image files into intricate tables. I'm no programmer (I know nothing about working with data structures, etc.), but I always thought the script would be really useful if it could analyze the image to determine where rowspans and colspans were needed (i.e., if cell 4 of row 1 and cell 4 of row 2 are the same color, make the rowspan of the first cell 2 and don't print the second cell... it seems easy enough on the surface, but it has to also consider rowspans and colspans of values greater than 2) to create optimized code.

I'll post the script if I can find it, but basically XPM is just a text file bitmap and a key. A simplified example:

Red=A
Black=B

BBB
BAB
BBB

Would represent a 3-pixel black square with a 1-pixel red dot in the middle.

Re:OS X

[Galvatron]

Why OS X on x86? Competition! Freedom! As it stands, if you want to switch to OS X, you have to buy a new computer, and lock yourself into a single hardware vendor. If Apple pulls some stunt down the road (like charging outrageous OS upgrade fees), you have to buy a new computer to switch back (or run some Free Unix variant). If the Mac OS ran on x86 hardware, we would finally have true competition in the OS arena, which can only mean good things for customers. If OS X for x86 made it easy to dual boot, I bet many people would give it a shot; they can always go back to windows if they don't like it.

Re:Agreed, DJVU is awesome... put in Mozilla pleas

[Gerv]

Of course to be truly useful as a JPEG replacement it needs to be included in browsers. People don't like having to mess with plugins.

Well, it'll never be in Mozilla/Netscape if things stay the way they are - there are two patents which apply to it, and you can only have a license for them if you are writing software that's purely GPLed. Read their licensing page.

Gerv

DJVU is patent-encumbered and partly closed

[Gerv]

From their website:

The Reference Library contains the entire DjVu decoder and much of the encoder, but it does not contain the sophisticated encoding strategies necessary for reaching the highest compression ratios. Among other things, the DjVu Reference Library contains the full DjVu decoder, the full IW44 wavelet encoder/decoder for continuous tone images, and the back-end of the JB2 bi-level image compression technique. It does not contain the code for separating document images into foreground and background layers, nor does it contain the code of the lossy JB2 scheme that achieves the highest compression ratios on bitonal images.

And...

Two patents apply to two very specific aspects of DjVu and DjVuLibre (described below). Those patents are owned by AT&T, but LizardTech has very broad rights to them and grants free and permanent licenses to them for the purpose of building GPL'ed software with the DjVu open source release.

So, DJVU could not appear in Mozilla/Netscape, because they wouldn't be granted a license to the patents. Mozilla is MPL/LGPL/GPL tri-licensed.

Gerv

Re:OS X

[GutBomb]

they could just cripple the parts of the system that are not open. there are more closed parts of OS X than open ones.

Understanding...

[sl3xd]

I think there was a bit of a misunderstanding.

I believe the poster meant that there are 2 different Itaniums:

One that is x86 backwards compatible (and only x86 backwards compatible)

One that is PA-RISC backwards compatible (and only PA-RISC backwards compatible)
And the poster thought it not much of a stretch to create a third version, that is PowerPC compatible.

Of course, there is only one Itanium core, and it handles all 3 (as you said). However, most RISC chips (such as the PA-RISC and PowerPC) at least have enough similarities that emulating PPC on PA-RISC (or using the PA-RISC decoder) is relatively simple; the opcodes may be different, but otherwise almost everything translates over directly.

eg. (example-- the actual binary is probably different)
Function to perform: A+B=C

"ADD A, B, C"='AF0F32BFh' in PPC machine language.

"ADD C, A, B"='CBBF0F32h' in PA-RISC machine language.
The difference is the opcode byte (AF v. CB), and ordering (A+B=C v. C=A+B)

The commands translate directly over, and only the formatting of the instruction matters. Easy emulation. x86 emulation is more of a bear: a single instruction can do different things, depending on the context (almost like operator overloading in assembly)

There have been similar rumors about using AMD chips; they go along these lines:

AMD Athlon & Opteron processors are really two processors: 1.) An x86 decoder, which translates the x86 instructions to 2.) AMD's completely original RISC core; each is roughly 1/2 of the total die size.

Take the upcoming Opteron, chop off the x86 decoder (which is about 1/2 of the chip), and use its RISC core natively (and emulate PPC)

Take the Opteron, and replace the x86 decoder with a PPC decoder (which would still be a smaller die than the x86 Opteron)

AMD is more likely to modify their design than Intel is.
Of course, the argument can be made 'why modify anything?'

As the poster said: x86 is on its last legs. The Opteron is likely the bed it will die in. There's really no reason to even have a CISC chip now that compiled languages are used instead of assembly.

There aren't many compelling things that show that VLIW is a better design paradigm than RISC. Few convincing reasons that VLIW (Itanium) is better than RISC (PowerPC)

Even Intel will have to debunk the MHz myth when trying to convince the public to buy the consumer version of Itanium, rather than the x86 Opteron.

Itanium and PowerPC have roughly equivalent SPEC scores at the same clock speeds.

There's not much to show that PowerPC is 'showing its age', as many of Itanium's touters claim. (It's more of a VLIW vs. RISC argument)

Apple has already done the processor emulation: When it moved from 680x0 to PowerPC. It's not as big a problem for them, having learned how to do it)

Re:Understanding...

[sl3xd]

As I said- I don't know much about VLIW at the current time; And I am quite ignorant of TransMeta's Code Morphing.

So I'll decline to comment further.

Re:OSX on x86

[sl3xd]

OSX will never be x86, nor will it work on 'commodity' (ie. non-Apple) hardware, Itanium/Opteron or not.

x86 is dying. Apple isn't known for living in the past.

Just because it's Intel doesn't mean it's x86.

Releasing OSX for x86 is completely moronic. Apple is a computer company, not a software company. They sell computers first, software second. If OSX ran on 'open' PC hardware, nobody would buy Apple computers-- they'd buy cheap hardware and OSX.

This is exactly what happened circa 1995 when there were Mac clones. The clones bled Apple dry. Steve Jobs saved Apple by making it a closed system again. Openness only works in a world that believes in openness. The clones exploited Apple's generosity, and it nearly killed Apple.

Pit any software against Microsoft, and expect Microsoft to attempt to kill it. Apple is doing well because they cooperate with Microsoft. If OSX were released for commodity PC hardware, and Microsoft will dump Office/Mac, and basically shut OSX out of the market (as it did with Netscape).

Free software is surviving Microsoft because it can't out-compete with Free Software's price. There's no company to bankrupt, and the software is largely donated by generous coders. Apple has no such protection. They can go bankrupt, and they don't have the hordes of programmers donating code that Free Software enjoys.

It does make sense

[tkrotchko]

Just because MacOS could be ported to run on X86 doesn't mean it would work on PC Compatible hardware.

A "Mac BIOS" that could make the Mac as closed and proprietary as it is today. The could make the same nice hardware, with the tight integration between OS and hardware that Mac customers expect. Only the processor inside might say Intel or AMD.

I think it would be a smart move on Apple's part.

No OS X for x86 - Here's Steve's FULL Quote...

[blakespot]

As taken from MacRumors.com:

• The quote regarding Steve Jobs on Intel may have been misrepresented by the original article. The quote came from Apple's Q3 Financial Analyst Meeting Q&A. From that broadcast, here is the entire quote in context:
  
  (5m 40s) Steve Jobs was asked about porting Mac OS X to Intel:
  
  Steve Jobs: "The roadmap on the PowerPC actually looks pretty good and there are some advantages to it. As an example, the PowerPC has something in it called AltiVec, we call the Velocity Engine -- it's a vector engine -- it dramatically accelerates media, much better than, as an example, the Intel processors or the AMD processors... so we actually eek out a fair amount of performance from these things when all is said and done. And the roadmap looks pretty good. Now, as you point out, once our transition to Mac OS 10 is complete, which I expect will be around the end of this year or sometime early next year and we get the top 20% of our installed base running 10, and I think the next 20 will come very rapidly after that. Then we'll have options, then we'll have options and we like to have options. But right now, between Motorola and IBM, the roadmap looks pretty decent. " ... so it looks like discussions of OS X on Intel/AMD may be a premature. (despite repeated speculation on this

A bit clearer, is his intent when taken in context.

blakespot

And if you subscribe right now...

[jejones]

Look more closely at the FAQ (under "Customer Services"). That $69.95 is a "special limited-time annual cost"...so later on, you get to pay more to watch commercials.

Re:"I have something to say..."

[dschuetz]

> Good god. Are you all 12???? Try The Who!!

Well, give some of us credit, remembering Neil Young, okay?

And, for the life of me, I can't place the Who song you're referring to. I've still got "Pink Moon" stuck in my head.

OSX on X86,the REAL FULL INCONTEXT QUOTE!!!!!!!!!!

[paradesign]

from macrumors.com, posted to alleviate questions of OSX on X86.

The quote came from Apple's Q3 Financial Analyst Meeting Q&A. From that broadcast, here is the entire quote in context:

(5m 40s) Steve Jobs was asked about porting Mac OS X to Intel:

Steve Jobs: "The roadmap on the PowerPC actually looks pretty good and there are some advantages to it. As an example, the PowerPC has something in it called AltiVec, we call the Velocity Engine -- it's a vector engine -- it dramatically accelerates media, much better than, as an example, the Intel processors or the AMD processors... so we actually eek out a fair amount of performance from these things when all is said and done. And the roadmap looks pretty good. Now, as you point out, once our transition to Mac OS 10 is complete, which I expect will be around the end of this year or sometime early next year and we get the top 20% of our installed base running 10, and I think the next 20 will come very rapidly after that. Then we'll have options, then we'll have options and we like to have options. But right now, between Motorola and IBM, the roadmap looks pretty decent. " ... so it looks like discussions of OS X on Intel/AMD may be a premature. (despite repeated speculation on this topic)

Ad Critic should change their subscription model

[BigJimSlade]

Have they considered supporting the site with ad revenue? Oh wait...

Hrm

[autopr0n]

Well, I don't know how easy it would be to yank off the x86 bit of the core and replace it with a PPC one. I'm sure the designs are pretty tightly coupled, for example the number of registers as well as things like SSE and MMX would need to work the way it does on an x86 chip.

I also seriously doubt the decoder takes up half the CPU space.

I also don't think Intel plans to sell the Itanium to the public for a long time. They may also add 64 bit capabilities to their Pentium line of chips to compete with AMD that way.

Re:Hrm

[sl3xd]

Well, I don't know how easy it would be to yank off the x86 bit of the core and replace it with a PPC one. I'm sure the designs are pretty tightly coupled, for example the number of registers as well as things like SSE and MMX would need to work the way it does on an x86 chip.

Not as coupled as you may think. SSE, MMX, 3dNow!-- all are either specialized floating point calculations (such as 128 bit floats-- but you can pack 2 64 bit floats in it and get a speed boost there), or are used for matrix operations. All of the above are decoded into many RISCOps, which are are then passed to the somewhat more 'generic' FP pipeline(s).

I also seriously doubt the decoder takes up half the CPU space.

You'd be greatly suprised. The whole RISC vs. CISC debate centers around this fact. The reason DEC dropped its VAX (CISC) processors and created the Alpha (RISC) was because the decode stage in the VAX would take more than 1/2 the design. Motorola dumped its 680x0 for PowerPC for the same reason. As did IBM and HP for their RISC chips. And Sun's had its RISC SPARC processors as long as I can remember.

Only Intel's x86 (and x86 clones from AMD, Cyrix, and others) remained CISC. Many analysts thought remaining CISC would kill Intel (as CISC chips are more expensive and more difficult to design). Only the sheer volume of x86 allows Intel to spend the money to develop CISC chips that perform as well as the much simpler RISC designs. (ie. if you look at what Intel spends to develop their x86 designs, and compare it to the development cost for a SPARC, Intel spends a lot more for a slower product).

AMD kept competitive by joining two chips: The decoder and the processing core. The decoder would only have to change if new instructions were added (such as SSE/2). And AMD could concentrate on a wickedly-fast RISC core.

The reason CISC ships have such a huge decoder is because of the complexity of the instructions.

A RISC instruction does only one task, in one way.

A RISC design has an absolute minimum of instructions, with no redundancy.

And to perform more complex tasks, you have to combine multiple instructions.

RISC instruction sets are not as easy to program in assembler with. (or, to be more accurate, it's a lot more tedious)

A CISC (like x86) has multiple-purpose, multiple method instructions. (somewhat like operator overloading in an Object Oriented language).

A CISC chip has ~300+ instructions, where a RISC chip has ~70. (With some RISC chips having as few as 40)

CISC assembler closely resembles a more high-level language.

Implementing high-level programming instructions in hardware takes a lot of transistors.

RISC designs have both less complicated (2-3 times simpler) instructions, and fewer instructions (by 3-6 times fewer).

Even in a RISC design, decode is at least 15% of the total design (and more if you use out-of-order execution).

I also don't think Intel plans to sell the Itanium to the public for a long time. They may also add 64 bit capabilities to their Pentium line of chips to compete with AMD that way.

Their marketing is a bit wishy-washy about that; I think they'll wait and see for the time being. There is a project at Intel called 'Yamhill' which is intended to be a 64-bit x86 clone (like AMD's Opteron).

The thing is there's not too much of a reason to move to 64-bit, save it be addressing more memory, or other things that require large integers. And if you're addressing that much memory, x86 is a lousy choice to start with.

Apple ad parody

[SuperMacNinja]

Gamer's take on Apple switch campaign.

Re:Motorola is a dead end.

[nobodyman]

re-read my post, zealot. I'm not talking speed, I'm talking increase in speed over time.

I've written assembly for 6800, x86, and MIPS. I know a thing or two about how processors handle instructions.

Re:Yeah, but..

[Hallow]

To get IE to support PNG alpha transparency you have to use pass the PNG through a DirectX filter, and it's pretty darn nasty.

Documentation on how to do it is here:

http://msdn.microsoft.com/library/default.asp?ur l= /workshop/author/filter/reference/filters/AlphaIma geLoader.asp

I tried it out. It's pretty darn slow, not to mention ugly. The DirectX Filters can do a lot more too, like resizing images on the fly and such. But of course it's MS/IE/DirectX only.

At least IE for MacOS doesn't seem to have any problems with standard transparent png's.

Lest we forget the roots

[kwerle]

OSX is, essentially, OpenStep 6.0 + Aqua (window system) + Carbon (Mac Apps) on Apple hardware. OpenStep ran on all kinds of hardware, including Intel. Aqua is new, and I'm sure written with an eye toward porting. Carbon once ran (somewhat) on Intel (I forget the codename of that one), and certainly it's far more portable now that it is just a set of libraries and not 'the whole OS stack'.

Re:OSX on x86

[Lars T.]

Not that this would be bad per se. Why does a modern PC need to have the A20 gate?

concepting ?1*@??!

[MountainLogic]

Concepting? What world do you live in? That's as bad as Colin Powell said yesterday about "operationalizing" something. /signed/ your former english teacher

Re:Agreed, DJVU is awesome... put in Mozilla pleas

[t]

DJVU will never get anywhere, they are far too restrictive. Just try and find software to create djvu files. The decoder is easy, they distribute a qpl version, works great, I viewed Shannon's original Information Theory paper with it. But ultimately, it will not be able to compete with JPEG2000 since performance et al will come second to easy access to working implementations.

See a comparison of djvu/jp2k etc...

t.

But $70???

[zCyl]

And I'm sure that there are marketing students in undergraduate and graduate programs nationwide that might find it a useful tool in doing research.

I'm sure they might, except that it costs $70 for them to even look at the ads to see if they want to use them. That's prohibitively expensive for any student to pay for an assignment.

At that annual fee, AdCritic has positioned itself to only be useful to people who are getting paid to look at ads. It's a niche, but a small one.

I hope eventually ad archives are more accessible to others, since they are often entertaining, and sometimes educational (as cultural indicators).

Already been proven

[bill_mcgonigle]

OpenStep developers almost always made these fat Mach-O binaries available. With the Apple toolchain , it's all automatic, the binary code is small compared to most of the other project resources, (NIB's images, etc.), and people now purchase computers with ridiculously large harddrives. They'll never notice a couple hundred K of extra binary code.

I was traumatized!

[autopr0n]

Actually, I have nothing against Macs, it's Apple I don't like :). I'm sure they're fine machines with OSX now, but I really doubt they are faster then price equivalent PCs. I've always found Apples advertising to be basically FUD and insulting to the intelligence of PC users. (as well as the people who they are advertising to). This seems to lead to lots of ignorant and shrill Mac advocates who insult and deride the PC. And yes, believe it or not some people like the PC. When you just dis it like that you dis those people.

So whenever I see rabid pro-Mac posts like these I usually will take the time to argue with them. Mac Advocates (and apples advertising department) should read the Linux Advocacy HOWTO sometime.

Re:OSX on x86

[sl3xd]

That's what people said when there were mac clones. Mac clones are what almost killed Apple. Apple is a computer company-- only a tiny fraction of their total income is from software, and most of their software profits come from QuickTime-- not OS X.

And dreaming that Apple's market share would grow is just that -- dreaming. Microsoft's strength is Windows-compatibility. And don't forget that since Microsoft owns a significant percentage of Apple, they still get to have their say.

And, finally -- why did Apple sell so much of itself to Microsoft? To get Office ported to OS X quickly. Apple isn't stupid. They know to be successful in the mass market, they have to have Microsoft Office, because that's what everybody uses at work (or school, etc).

And if you think Apple cares about your $129, you're kidding yourself. They want the $800-3000 profit they recieve from selling hardware. They can stay in business (and are doing better than many of the clones makers, like Gateway). Apple cannot stay in business if there are clones running Macintosh.

God, just look at a picture

[autopr0n]

See this this image of an Athlon CPU? (on this page. if the have hotlink protection)See the block labeled 'MENG/EDEC'? That's the decoder. It can do 3 at a time and as you can see it does not take up half the CPU.

also, CISC != high level langauge.

[autopr0n]

CISC assembly is not anything like any high level language I've ever seen, and I've seen a lot. Any trained monkey would put an x86 program in a pile with other Assembly stuff and not in a pile with other high level stuff like C or java or Scheme.

You missed some of it.

[sl3xd]

There's a lot more to the decode than that.

Let me get this out of the way: I design microprocessors (but mine are more specialized than the generic CPU's pumped out by Intel & AMD). In every case it's much easier, and less costly to use a RISC design. And the RISC design crunches numbers just as quickly.

I'll admit this: I neglected cache memory and cache control circuitry. This takes up about 1/3 of the design, and modifies things a bit-- but the relative size of the execution pipelines to the instruction decode is still roughly 50/50.

Instruction decode also requires the following (and possibly more) sections (That you appear to have missed):

PreDecode Array/CTL

Scan/Align

Instruction Control Unit

Branch Prediction

Floating Point Control

Floating Point Scheduler

And, the Athlon can not decode 3 instructions at the same time. It can dispatch 3 instructions to the integer RISC pipelines at the same time. This is an important distinction, as 1 CISC instruction is equivalent of 1 to 10 (or more) RISC instructions. Each RISC instruction takes 1 clock tick to execute. But a CISC operation can take anywhere from one to several hundred clock cycles to execute. Overall, it takes the same amount of time-- its' just that the RISC does it in several small steps to 1 huge CISC step.

In other words, it decodes 1 x86 instruction, and then can dispatch up to 3 of the 'RISC ops' that are required to do the same task as 1 x86 instruction. (But remember-- 1 x86 instruction may be a lot more than just 3 'RISC ops').

Another thing to remember is that the size of the instruction decode is relatively static-- but the execution stage has grown progressively larger (by adding multiple identical pipelines). Only now when there are 5+ pipelines is the execution stage approaching the size of the decoder.

With fewer pipelines (such as '486, which has 1 FP and 1 integer, which aren't interchangeable), the decode was much larger than the execution logic.

In contrast, a RISC design (single-pipeline) takes up ~15% of the total core size. Superscalar (multiple identical pipelines), and adding cache can reduce the decode to 2-3% of the total die.

Where a CISC design (single pipeline) the decode is ~75% (this is textbook data; from an x86 architecture text). Superscaler, cached designs reduce the decode to ~33% of the total die.

So there's a RISC decode (2-3% of total die), vs. CISC decode (33% of total die)

Don't try to tell me that the difference is small or insignificant. The PowerPC G4 (which SPEC's faster than an Athlon of twice its clock speed) has less than half the transistors of the Athlon. And it requires less than 1/3 of the power.

Hrm...

[autopr0n]

Well, I think we are thinking of diffrent things, when I say "decode" I'm just talking about the conversion from CISC to RISC, not from general CPU instructions to microcode. Do you have a link to the SPEC benchmarks for PPC and Athlon?

Re:Hrm...

[sl3xd]

If that's the case, we most definately are.

And, one other point: Only CISC uses microcode (as using dedicated logic to decode is even more complex). RISC uses pure dedicated logic to decode. Using a microcode decoder does reduce complexity of a CISC design (a dedicated logic decoder is trulyl monstrous in a CISC design). RISC is able to get away with using dedicated decoding because they have so few (and considerably more simple) instructions.

And microcode has made huge leaps in efficiency lately -- current designs have microcode that executes close to 95% as fast as pure dedicated logic (so it's only about 5% slower to decode).

And for a link (I didn't try hard) you can do a quick 1-min google search...

But here's some I turned up:

http://www.geocities.com/imac_driver/cpu.html (not SPEC, but it does compare FLOPS (Floating Point Ops/sec) and has links to other comparisons.

Re:OS X

[MoneyT]

So how is that different than what they have now? All the other hardware is PC hardware