Apart from all the MAC/Mac jokes, I'm wondering what significance (if any) this has for the future of the AIM client on Mac OS X. The last time the official AIM client for Mac was updated was (according to its download page) on February 18, 2004, i.e. over two years ago. Considering that most people who want to develop an AIM client for the Mac have already done so by using the GAIM core libraries (Adium X being one chief example), what exactly is making an SDK with a small pile of licensing restrictions (you are "not permitted to build Custom Clients that are multi-headed or interoperable with any other IM network" ? Wonder why that is...[/sarcasm]) going to solve?
I appreciate the gesture, but I think a lot more people would benefit if they'd spend their time fixing up the official AIM client for people who don't particularly like iChat or any of the current F/OSS alternatives rather than releasing an SDK that probably won't spur many/any more F/OSS clients because developers will find its license a bit too restrictive.
Well, it may just be a mistake... but it's rather odd that if you look at the Darwin source code made available on Apple's site, you'll find the Darwin kernel ("xnu") is present in the PPC source for both OS X 10.4.4 and 10.4.5, but is missing from the x86 source for both of those. I highly doubt the two kernels are built from identical source code. From what I've read, some amount of TPM and/or EFI dependency is built into the OS X/Intel kernel itself, and that fellow-whose-name-it-is-apparently-a-DMCA-violatio n-to-mention had to heavily modify the kernel to get OS X booting on generic hardware. If the kernel source was identical between PPC and x86, a freshly compiled kernel from the Darwin source code should have made for a fairly simple drop-in replacement job, but it didn't.
It's not a smoking gun of course (and I'm no expert on the architecture of OS X), but it does look to me like Apple is withholding Darwin source code in an attempt to frustrate OS X hacking efforts.
"Qemu may not run as fast as vmware does now but it's here, it's free and you can change whatever you want about it. The same is not true for vmware"
Well, to be entirely fair, that's because they do two different things. Qemu does CPU emulation, VMWare does virtualization. There is a module for QEMU that implements direct code execution (on x86 at least), but as per its webpage here, it is a "free to use, but it is a closed source proprietary product," so you can't exactly 'change whatever you want about it'.
It would be a shame to lose them to the ravages of time because of insane copyright laws, like what can (and is happening) to film from 1923 on.
Right, and I'd like to take a moment to point out what I think is the single most important aspect of this project. To quote their copyright policy page here:
The raw transfers created by the University of California are in the public domain.
In my opinion, one of the greatest things crippling the public domain today is the fact that even when public domain sources exist, they're inaccessible to those who might want to use them. I ran into this problem myself some months back when researching some vintage Broadway material for a recording project. Musicals that were written by people like George M. Cohan circa 1905 are clearly now public domain and, in theory, may be performed, recorded etc. at will. In practice, however, most of the original scripts, musical scores etc. have been bought up by firms like Tams-Witmark (who are happy to license you their own copyrighted reprints, for the right price...) or are stowed away in private collections. The net result is that it is very difficult to actually take advantage of the public domain status of these works because almost no verifiably public domain sources are actually available to the general public.
The solution is not necessarily to use "copyleft"-style licensing to guarantee reprints, restorations, etc. of public domain works remain open. That places restrictions (granted pro-copying ones) on the use of the works, and the original intent of the public domain was that after a certain time no such restrictions should exist. If Tams-Witmark spends the time, effort, and money "polishing up" a century-old Broadway score with new typesetting and the like, then as far as I'm concerned they should be able to offer that for sale and make whatever profit they can with it. What I *do* object to is the fact that the limited availability of original public domain sources effectively makes licensing of a copyrighted derivative the *only* reasonable way to obtain some older works.
In short, I say UCSB has done The Right Thing(tm) here. They've chosen their own terms (in this case a Creative Commons License) for their restored derivatives, but they've explicitly granted the raw transfers to the public domain. The net effect is as though there were thousands more of the original cylinders to work from, available to all those who want one, and the monopoly-of-sorts I described above does not occur. We need more online archives of all manner of public domain material, and it's (in my opinion) particularly vital that the copies offered online be public domain themselves. It's possibly the best way we have of preserving this culture for the future.
I thought this horse had already been beaten to death... but there are at least two good reasons Apple will probably never ship iTunes for Linux:
1)Technical - iTunes depends on QuickTime (the iTunes installer for Windows even includes it...), which ran on Windows for years and years before iTunes even existed. Not that porting iTunes to Windows was necessarily easy, mind you... but at least they had the groundwork in place before they got started. There's no native QuickTime for Linux (yet?), so getting that running would be another item on the to-do list before an iTunes Linux port could be crafted. Add on all the complexity of choosing which Linux distro(s) to support and working around their differences (keeping in mind that even on Windows, iTunes only attempts to support the 2000/XP family, not 9x...) and you get a fairly tricky problem to be solved just to gain what is still a fairly small market.
2)Political - In the old Apple slogan "Think Different," *they* are the "Different." For years, Apple has been preaching that if you want to escape from Microsoft et. al., *they* are your alternative. Don't like Windows? Buy a Mac... that simple. Except that it's not that simple anymore... Linux is getting more mature, refined, and competitive all the time and is in some ways already a bigger threat to Microsoft than Apple is now (or probably ever will be again). If Apple ships iTunes for Linux, it will seem as though they're acknowledging that Linux is ready for desktop/home use, and that there are ways to "Think Different" that don't involve Apple.
Apple might have an iTunes for Linux someday. Far stranger things have already happened (Intel switch, anyone?), and they have the resources to overcome the hurdles in point #1 if they want to. The only question is: given point #2, do they *really* want to?
Well, remember the articles only refer to hacking the special build of OS X 10.4.1 provided with the Developer Transition Kits to run on generic PC hardware. This isn't a shipping or release version of Mac OS X, just a developer testing tool. Apple makes it crystal clear in the Terms & Conditions you have to agree to in order to get a kit that they are not obligated to provide software updates for them (or much other tech support for that matter, which is one of the reasons I didn't bother ordering one of the kits even though I'm an ADC member [the other being that, as the few software projects I release publicly are all freeware or F/OSS stuff, paying $999 to rent the thing is a bit of a non-starter for me]).
Back to my main point... if by "those users," you mean the group of people running this build of OS X on PCs right now... I don't think they need to worry about a security patch from Apple screwing things up, because even registered developers aren't getting updates for this release of OS X from what I understand. I assume one big reason for that is just what we're discussing here... leaked releases. Apple surely knew it was only a matter of time before somebody got the developer OS X for Intel running on generic PCs, so they want to "reduce the shelf life," if you will, of this OS as much as possible. If they don't release any security patches or other updates... well that helps to limit the period for which the OS is useful to anyone (keeping in mind all developers are going to have to return their test systems and buy production Intel Macs by the end of next year anyway).
As for the bigger question (i.e. if/when the retail/supported & updated OS X for Intel is hacked and run on generic PCs), I suspect Apple will make it a real minefield for unauthorized users trying to get patches. Unfortunately that could also leave room for a flaw in their validation checks to frustrate users of genuine Apple hardware, but we'll just have to see how that one pans out.
"I'll be so glad when you vinylphiles finally all die out.
Hey, I tease.:)
But seriously... Nyquist and all that."
Nyquist and all that? All that the Nyquist theorem says on the subject is that a sampling audio system like PCM should, in theory, be able to reproduce signals with frequencies up to 1/2 of the sampling rate faithfully. But in the real world, there are at least two problems with that:
1) The low-pass filters used on the signal path are physical devices, not theoretical concepts. As such, they can't be absolutely perfect... they introduce phase distortions and begin attenuating at frequencies somewhat lower than 1/2 Fs.
2) Even if the filters were "perfect" (not attenuating or introducing phase distortion until 1/2 Fs, at which point the attenuation becomes infinite)... well, the jury is still out on whether 22050hz (the theoretical upper bound given the 44.1khz sampling rate of CDs) is really high enough. There's some evidence to suggest that even if we can't "hear" frequencies above 22.050khz, they can have an effect on the way we perceive lower frequencies that we can hear.
Just to be fair to both sides of the argument though...
"CDs are still far worse sounding than vinyl."...only on excellent vinyl playback equipment. It tends to be tougher to produce a mechanical device like a turntable cartridge with the same level of consistency that can be expected when producing ICs and the like. That (along with other factors like simple supply and demand) is why decent vinyl playback stuff tends to be quite a bit more expensive than decent CD players do. I have a reasonably high-end turntable and I enjoy using it tremendously... but I have to admit it wasn't cheap compared to digital gear in its league.
To return to the digital downloads aspect of the article a bit though... I have to completely agree with the poster who shuns download services for poor quality. The only times I've extensively used iTunes were the Pepsi free song promotions, and if I found any songs I really really liked... well I went on Amazon or to my local record store and sought out the CDs to re-rip as DRM-free Apple Lossless. Better sound quality and the ability to use the format of my choice will make CDs the clear winner in my book for a long time to come.
The parent is entirely right, especially if the "professionals" in question are in one of Apple's core markets: video and/or sound production. People in those fields who use their Macs every day to get jobs done for clients are often a bit hesitant to upgrade, especially for something as major as a CPU architecture switch.
To this day, years after Steve Jobs went on stage and proclaimed that OS 9 was dead, you'd probably be a little hard pressed to find any post-production house in LA or NYC that didn't have at least one old workhorse Mac, probably a 450-500mhz G4 or thereabouts, still running OS 9.2 and still being used to get work done. Are most of the applications they're using now available in OS X versions? Probably, although not all of them are. But to use the old saying, "if it ain't broke, don't fix it."
In fact, a lot of professionals still use machines a lot older than that on a daily basis. A few months back I was in a recording studio in Pittsburgh where, right alongside a state-of-the-art network of XServe G5s and PowerMac G5s running the very latest version of Pro Tools, they still kept and used an old Quadra 700 to run Sound Designer II on an old version of System 7. As we sit around talking about the PPC->Intel transition, they still found frequent enough use for an over 12 year old machine with a 25mhz 68040 chip to keep it out alongside the latest-and-greatest rather than putting it in the storage closet.
In conclusion, lots of people in Apple's core markets will still buy the remaining PPC Macs and use them for years to come, just because they'll get the job done.
... I would expect that it would be a lot like the old Rhapsody DR1 and DR2 releases were on x86 (anybody else remember those things?), except with newer and shinier "eye candy." That is to say, assuming it can even be booted, hardware support will be *extremely* limited. In fact, it may be even worse because while the Rhapsody DR releases on x86 were intended to target beige-box PCs (if only a few models thereof), this build of Mac OS X was only intended to target a single, very specific "PC"... namely the transition kit system Apple is making.
I would hazard a guess that would-be Intel Mac "owners" will have the most luck with something as similar to the transition system as reasonably possible (which would I believe be a Pentium 4 660, running on an Intel OEM motherboard w/ one of the 900-series chipsets and using its onboard graphics, maybe an Apple-supported optical drive couldn't hurt, etc...). And that's assuming Apple hasn't had special tweaks made to the BIOS that OS X/Intel looks for before running (they lock out non-Apple optical drives with special firmware, why would they be so careless when talking about a full OS running on non-Apple-blessed hardware, unless of course this is really a marketing conspiracy...?).
In summary, I think most of the people who download this (again, assuming it actually has leaked at all...) will be "eleet dood skript kiddiez" who expect they'll be able to pop a DVD into their Athlon 64 PC w/ killer nVidia graphics card, boot up and install OS X, and get a free "SuperDuperPowerMac G6+!!!" to play with iMovie and such on. But the chances of it actually being that simple are slim-to-none. It could still be a fun toy for more knowledgeable computer geeks though.
There's also a few Safari bugs that snuck in late in the cycle that haven't been addressed yet
A few Safari bugs, yes... also there's apparently something about syncing Address Book data with.Mac that was left unfixed. It was listed as a "Known Issue" when this build (8B15) was given to developers last week (I suppose it's fair game to comment on now that the release is public). I've never experienced the issue myself, but I was sort of expecting them to take a bit more time to fix it, and release build 8B17 (or whatever) with a fix included as the public 10.4.1. Apparently they've decided either a) the bug is something they can fix server-side w/.Mac, or b) it's something sufficiently rare or minor to wait for 10.4.2. I'd probably bet on it being the second one.
So, as you said, it's "right on schedule." I daresay this update being released just now is mainly about getting over the psychological barrier some Mac users tend to have about upgrading to a "10.x.0" release, and also to tidy up some loose ends for Final Cut Studio (which, if memory serves, should ship soon and originally listed 10.4.1 as a requirement until DVD Player 4.6 came out separately). I realize they can't fix everything in one go, but it could be that this release came just a bit early to meet the deadline imposed by Final Cut Studio.
I couldn't agree more... I'd love to see lossless music on iTMS too. But I'm fairly certain they'll never do it because of the old burn-and-re-rip DRM hole.
As it stands now, if you burn your 128k AAC purchase from iTMS to a CD and re-rip the result to strip off the FairPlay DRM in an Apple-sanctioned manner, you've either a) lost some quality along the way by using another lossy format to re-encode, or b) grossly oversized the file by using Apple Lossless to re-encode the previously lossy material. But if they offer lossless tunes for download, then that same process will result in a perfect DRM-less copy (unless of course faulty hardware or something similar caused a bit or two to get lost along the way).
They'd never be able to sell the record companies on that one. So, unfortunately, I doubt we'll ever see lossless downloads from iTMS unless they prevented them from ever being burned to CD (in which case a lot of the desirability flies right out the window).
PCI-E isn't really necessary - you'd be hard-pressed to fully use the available bandwidth on graphics cards - AGP 8x is damn fast.
I directly addressed this in my post, right after the part where you conveniently stopped quoting me. Yes, AGP 8x is "damn fast." No, current graphics chipsets don't really need all of its bandwidth, let alone that which would be provided by PCI-E. But as I said originally, it's "a performance gap (real or perceived, it matters not in the marketplace)." Even disregarding that, the sad fact is that PCs far outnumber Macs in the world, and their transition to PCI-E is well underway. Once that's done, graphics card makers like ATI or nVidia will be even more reluctant to sink money into Mac graphics card development. It's a small market now, and in the future it will be a small market that is even more expensive to develop for because of the need to tie new graphics chips to the older AGP 8x bus that PCs no longer use. Graphics cards for Mac could become even more expensive than they are now, the delay between introductions of the latest chips for PCs and Macs could become even longer than it is now, and neither of those are good signs for a machine that is meant to "likely last 5+ years" as you said. A machine that is meant to last a long time needs to be designed with an eye toward future expandability, not just current needs.
Seriously, what would you add to make the G5 more enticing? it's arguably faster than any dell available, and will likely last 5+ years, so what's missing?
Enticing to new buyers? Nothing. It's great. Enticing to current owners of Macs, or to power users of PCs looking to switch? Only a few things, but I already named them in some detail in my original post.
We refresh our product lines roughly once every nine months. We've been doing it that way for years now. Why is this a surprise?
Not to be rude, but I don't recall ever saying it was a surprise...
Hopefully nothing. "Power Macintosh G5" is a brand item for us. We don't want to release a product and then suddenly drop it. Instead, we want to release a product and maintain it for several years
In the computer industry, "maintaining" a product for several years sometimes means updating components to be more current technology. How would offering faster RAM, for instance, damage the brand recognition of "Power Macintosh G5"? If anything, such a change would strengthen the idea that the PowerMac G5 is truly a world-class workstation computer and provides the best performance available. Saying that "hopefully nothing" changed in the G5 in nearly two years makes it sound like Apple wants a computer to be like a piece of fine furniture. Design it once, change it never. But that's not a very realistic way for the power user market the G5 is apparently targeted towards to view a machine.
Bottom line: We don't just roll out whole new products willy-nilly. Part of what we sell our customers is stability. One of the things you know when you buy a Mac -- most of the time -- is that the thing you buy isn't going to be just totally lame next month.
Few people would call a logic board revision a "whole new product." Saying that the suggestions I made regarding PCI-Express or DDR2 would be rolling out a whole new product is like saying that the Sawtooth G4 was a "whole new product" because it had AGP instead of the PCI graphics like the prior Yikes! model had. Or that the G4's later move from PC133 SDRAM to DDR SDRAM was a whole new product. They were still called G4s, and they still sold well. I don't suppose the brand recognition of a "Power Macintosh G4" suffered for it. And in any event, because of the predictable 9-month upgrade cycles you mentioned, only a few unlucky buyers on the edge of an upgrade cycle would be faced with having a "totally lame" machine soon after their purchase.
My "Bottom line": Apple can produce and has produced more significant performance upgrades to their pro line than either of the G5 upgrades thus far has been. The real issue is that the G5 seems to be suffering from a premature onset of the sluggish development pace that characterized the G4 late in its run. And that's a worrying sign to Mac users like myself.
Actually, one thing I think is even more damaging to Apple than the "rumor mill" is the perceived stagnation of the PowerMac lineup. To an outsider (like a consumer in the market for a machine), what has changed in the G5 in nearly two years since its introduction?
1.CPU and FSB speeds have increased. 2.Bigger hard drives. 3.Faster Superdrives. 4.Faster video cards.
I've left out some details obviously, but those are a few major features someone might look at when considering a G5. Of the things I've listed, only the first item (faster CPUs) can't be readily added to either of the two previous generations of G5 at a very nominal cost and with little effort. So while the new PowerMacs are quite impressive machines and are certainly great for first-time buyers, what they may fail to do is to entice any current G5 owners to upgrade to a newer model. And considering that the first G5s appeared nearly two years ago now, that current-owner-looking-to-upgrade-soon market is going to be sizeable in the relatively near future.
What would entice one of those owners to buy a new G5? Perhaps moving to PCI-Express graphics. Perhaps moving to DDR2 RAM. Adopting any of those technologies that 1) the PC market already has, and thus yields a performance gap (real or perceived, it matters not in the marketplace), and 2) that can't be retrofitted onto an older G5 logic board for low cost and effort. Lacking dual-core CPUs wouldn't even be very disappointing if the rest of the architecture surrounding single-core chips got a bit of a revamp.
In short, *that* is the sort of upgrade that will capture both new and existing Mac users to buy a new G5, and the lack of such an upgrade is probably as big of an influence as the "rumor mill" you blamed in damaging Apple and keeping PowerMac sales a bit low. As a current G5 user, I for one hope such an update comes soon. My checkbook will be ready.
Actually, following the House bill's description on thomas.loc.gov leads to a page, showing that Sen. Hatch is the main sponsor of the Senate versions of this legislation.
I would guess that as Sen. Hatch is the one who originally introduced the bill to the Senate, he would vote in its favor when the time comes... assuming of course that he's not too offended by any changes made in the House.
I might mention that Kevin Gerich's widget set makes Firefox's HTML controls look much more presentable on Mac, in my opinion. It's not quite the same as having native Aqua widgets, but it's a start. Granted they aren't bundled with the application by default, nor do they solve any of the other OS integration issues you mentioned.
That having been said, I agree with the assessment that Firefox for Mac has a lot of catch-up to do to match Safari in terms of aesthetics. It's one of the biggest cons of choosing Firefox on the Mac platform. Safari, as Apple's own in-house effort, gets a level of fit-and-finish with the rest of the OS that third-party developers can have a tough time matching.
On the other hand, the biggest pro for Firefox on Mac (in my opinion) is the expandability. Safari doesn't have Adblock, BugMeNot, or any of my other favorite extensions. Even Camino doesn't support them. So in my case, I choose expandability over aesthetics and use Firefox as my default browser on Mac.
Ideally though, it would be possible to have both. Maybe in time and with further Firefox development.
Slashdot Mirror
Apart from all the MAC/Mac jokes, I'm wondering what significance (if any) this has for the future of the AIM client on Mac OS X. The last time the official AIM client for Mac was updated was (according to its download page) on February 18, 2004, i.e. over two years ago. Considering that most people who want to develop an AIM client for the Mac have already done so by using the GAIM core libraries (Adium X being one chief example), what exactly is making an SDK with a small pile of licensing restrictions (you are "not permitted to build Custom Clients that are multi-headed or interoperable with any other IM network" ? Wonder why that is...[/sarcasm]) going to solve?
I appreciate the gesture, but I think a lot more people would benefit if they'd spend their time fixing up the official AIM client for people who don't particularly like iChat or any of the current F/OSS alternatives rather than releasing an SDK that probably won't spur many/any more F/OSS clients because developers will find its license a bit too restrictive.
-Frank
Well, it may just be a mistake... but it's rather odd that if you look at the Darwin source code made available on Apple's site, you'll find the Darwin kernel ("xnu") is present in the PPC source for both OS X 10.4.4 and 10.4.5, but is missing from the x86 source for both of those. I highly doubt the two kernels are built from identical source code. From what I've read, some amount of TPM and/or EFI dependency is built into the OS X/Intel kernel itself, and that fellow-whose-name-it-is-apparently-a-DMCA-violatio n-to-mention had to heavily modify the kernel to get OS X booting on generic hardware. If the kernel source was identical between PPC and x86, a freshly compiled kernel from the Darwin source code should have made for a fairly simple drop-in replacement job, but it didn't.
It's not a smoking gun of course (and I'm no expert on the architecture of OS X), but it does look to me like Apple is withholding Darwin source code in an attempt to frustrate OS X hacking efforts.
-Frank
"Qemu may not run as fast as vmware does now but it's here, it's free and you can change whatever you want about it. The same is not true for vmware"
Well, to be entirely fair, that's because they do two different things. Qemu does CPU emulation, VMWare does virtualization. There is a module for QEMU that implements direct code execution (on x86 at least), but as per its webpage here, it is a "free to use, but it is a closed source proprietary product," so you can't exactly 'change whatever you want about it'.
-Frank
Right, and I'd like to take a moment to point out what I think is the single most important aspect of this project. To quote their copyright policy page here:
In my opinion, one of the greatest things crippling the public domain today is the fact that even when public domain sources exist, they're inaccessible to those who might want to use them. I ran into this problem myself some months back when researching some vintage Broadway material for a recording project. Musicals that were written by people like George M. Cohan circa 1905 are clearly now public domain and, in theory, may be performed, recorded etc. at will. In practice, however, most of the original scripts, musical scores etc. have been bought up by firms like Tams-Witmark (who are happy to license you their own copyrighted reprints, for the right price...) or are stowed away in private collections. The net result is that it is very difficult to actually take advantage of the public domain status of these works because almost no verifiably public domain sources are actually available to the general public.
The solution is not necessarily to use "copyleft"-style licensing to guarantee reprints, restorations, etc. of public domain works remain open. That places restrictions (granted pro-copying ones) on the use of the works, and the original intent of the public domain was that after a certain time no such restrictions should exist. If Tams-Witmark spends the time, effort, and money "polishing up" a century-old Broadway score with new typesetting and the like, then as far as I'm concerned they should be able to offer that for sale and make whatever profit they can with it. What I *do* object to is the fact that the limited availability of original public domain sources effectively makes licensing of a copyrighted derivative the *only* reasonable way to obtain some older works.
In short, I say UCSB has done The Right Thing(tm) here. They've chosen their own terms (in this case a Creative Commons License) for their restored derivatives, but they've explicitly granted the raw transfers to the public domain. The net effect is as though there were thousands more of the original cylinders to work from, available to all those who want one, and the monopoly-of-sorts I described above does not occur. We need more online archives of all manner of public domain material, and it's (in my opinion) particularly vital that the copies offered online be public domain themselves. It's possibly the best way we have of preserving this culture for the future.
-Frank
I thought this horse had already been beaten to death... but there are at least two good reasons Apple will probably never ship iTunes for Linux:
1)Technical - iTunes depends on QuickTime (the iTunes installer for Windows even includes it...), which ran on Windows for years and years before iTunes even existed. Not that porting iTunes to Windows was necessarily easy, mind you... but at least they had the groundwork in place before they got started. There's no native QuickTime for Linux (yet?), so getting that running would be another item on the to-do list before an iTunes Linux port could be crafted. Add on all the complexity of choosing which Linux distro(s) to support and working around their differences (keeping in mind that even on Windows, iTunes only attempts to support the 2000/XP family, not 9x...) and you get a fairly tricky problem to be solved just to gain what is still a fairly small market.
2)Political - In the old Apple slogan "Think Different," *they* are the "Different." For years, Apple has been preaching that if you want to escape from Microsoft et. al., *they* are your alternative. Don't like Windows? Buy a Mac... that simple. Except that it's not that simple anymore... Linux is getting more mature, refined, and competitive all the time and is in some ways already a bigger threat to Microsoft than Apple is now (or probably ever will be again). If Apple ships iTunes for Linux, it will seem as though they're acknowledging that Linux is ready for desktop/home use, and that there are ways to "Think Different" that don't involve Apple.
Apple might have an iTunes for Linux someday. Far stranger things have already happened (Intel switch, anyone?), and they have the resources to overcome the hurdles in point #1 if they want to. The only question is: given point #2, do they *really* want to?
-Frank
Well, remember the articles only refer to hacking the special build of OS X 10.4.1 provided with the Developer Transition Kits to run on generic PC hardware. This isn't a shipping or release version of Mac OS X, just a developer testing tool. Apple makes it crystal clear in the Terms & Conditions you have to agree to in order to get a kit that they are not obligated to provide software updates for them (or much other tech support for that matter, which is one of the reasons I didn't bother ordering one of the kits even though I'm an ADC member [the other being that, as the few software projects I release publicly are all freeware or F/OSS stuff, paying $999 to rent the thing is a bit of a non-starter for me]).
Back to my main point... if by "those users," you mean the group of people running this build of OS X on PCs right now... I don't think they need to worry about a security patch from Apple screwing things up, because even registered developers aren't getting updates for this release of OS X from what I understand. I assume one big reason for that is just what we're discussing here... leaked releases. Apple surely knew it was only a matter of time before somebody got the developer OS X for Intel running on generic PCs, so they want to "reduce the shelf life," if you will, of this OS as much as possible. If they don't release any security patches or other updates... well that helps to limit the period for which the OS is useful to anyone (keeping in mind all developers are going to have to return their test systems and buy production Intel Macs by the end of next year anyway).
As for the bigger question (i.e. if/when the retail/supported & updated OS X for Intel is hacked and run on generic PCs), I suspect Apple will make it a real minefield for unauthorized users trying to get patches. Unfortunately that could also leave room for a flaw in their validation checks to frustrate users of genuine Apple hardware, but we'll just have to see how that one pans out.
-Frank
"I'll be so glad when you vinylphiles finally all die out.
:)
...only on excellent vinyl playback equipment. It tends to be tougher to produce a mechanical device like a turntable cartridge with the same level of consistency that can be expected when producing ICs and the like. That (along with other factors like simple supply and demand) is why decent vinyl playback stuff tends to be quite a bit more expensive than decent CD players do. I have a reasonably high-end turntable and I enjoy using it tremendously... but I have to admit it wasn't cheap compared to digital gear in its league.
Hey, I tease.
But seriously... Nyquist and all that."
Nyquist and all that? All that the Nyquist theorem says on the subject is that a sampling audio system like PCM should, in theory, be able to reproduce signals with frequencies up to 1/2 of the sampling rate faithfully. But in the real world, there are at least two problems with that:
1) The low-pass filters used on the signal path are physical devices, not theoretical concepts. As such, they can't be absolutely perfect... they introduce phase distortions and begin attenuating at frequencies somewhat lower than 1/2 Fs.
2) Even if the filters were "perfect" (not attenuating or introducing phase distortion until 1/2 Fs, at which point the attenuation becomes infinite)... well, the jury is still out on whether 22050hz (the theoretical upper bound given the 44.1khz sampling rate of CDs) is really high enough. There's some evidence to suggest that even if we can't "hear" frequencies above 22.050khz, they can have an effect on the way we perceive lower frequencies that we can hear.
Just to be fair to both sides of the argument though...
"CDs are still far worse sounding than vinyl."
To return to the digital downloads aspect of the article a bit though... I have to completely agree with the poster who shuns download services for poor quality. The only times I've extensively used iTunes were the Pepsi free song promotions, and if I found any songs I really really liked... well I went on Amazon or to my local record store and sought out the CDs to re-rip as DRM-free Apple Lossless. Better sound quality and the ability to use the format of my choice will make CDs the clear winner in my book for a long time to come.
-Frank
The parent is entirely right, especially if the "professionals" in question are in one of Apple's core markets: video and/or sound production. People in those fields who use their Macs every day to get jobs done for clients are often a bit hesitant to upgrade, especially for something as major as a CPU architecture switch.
To this day, years after Steve Jobs went on stage and proclaimed that OS 9 was dead, you'd probably be a little hard pressed to find any post-production house in LA or NYC that didn't have at least one old workhorse Mac, probably a 450-500mhz G4 or thereabouts, still running OS 9.2 and still being used to get work done. Are most of the applications they're using now available in OS X versions? Probably, although not all of them are. But to use the old saying, "if it ain't broke, don't fix it."
In fact, a lot of professionals still use machines a lot older than that on a daily basis. A few months back I was in a recording studio in Pittsburgh where, right alongside a state-of-the-art network of XServe G5s and PowerMac G5s running the very latest version of Pro Tools, they still kept and used an old Quadra 700 to run Sound Designer II on an old version of System 7. As we sit around talking about the PPC->Intel transition, they still found frequent enough use for an over 12 year old machine with a 25mhz 68040 chip to keep it out alongside the latest-and-greatest rather than putting it in the storage closet.
In conclusion, lots of people in Apple's core markets will still buy the remaining PPC Macs and use them for years to come, just because they'll get the job done.
-Frank
... I would expect that it would be a lot like the old Rhapsody DR1 and DR2 releases were on x86 (anybody else remember those things?), except with newer and shinier "eye candy." That is to say, assuming it can even be booted, hardware support will be *extremely* limited. In fact, it may be even worse because while the Rhapsody DR releases on x86 were intended to target beige-box PCs (if only a few models thereof), this build of Mac OS X was only intended to target a single, very specific "PC"... namely the transition kit system Apple is making.
I would hazard a guess that would-be Intel Mac "owners" will have the most luck with something as similar to the transition system as reasonably possible (which would I believe be a Pentium 4 660, running on an Intel OEM motherboard w/ one of the 900-series chipsets and using its onboard graphics, maybe an Apple-supported optical drive couldn't hurt, etc...). And that's assuming Apple hasn't had special tweaks made to the BIOS that OS X/Intel looks for before running (they lock out non-Apple optical drives with special firmware, why would they be so careless when talking about a full OS running on non-Apple-blessed hardware, unless of course this is really a marketing conspiracy...?).
In summary, I think most of the people who download this (again, assuming it actually has leaked at all...) will be "eleet dood skript kiddiez" who expect they'll be able to pop a DVD into their Athlon 64 PC w/ killer nVidia graphics card, boot up and install OS X, and get a free "SuperDuperPowerMac G6+!!!" to play with iMovie and such on. But the chances of it actually being that simple are slim-to-none. It could still be a fun toy for more knowledgeable computer geeks though.
-Frank
There's also a few Safari bugs that snuck in late in the cycle that haven't been addressed yet
.Mac that was left unfixed. It was listed as a "Known Issue" when this build (8B15) was given to developers last week (I suppose it's fair game to comment on now that the release is public). I've never experienced the issue myself, but I was sort of expecting them to take a bit more time to fix it, and release build 8B17 (or whatever) with a fix included as the public 10.4.1. Apparently they've decided either a) the bug is something they can fix server-side w/ .Mac, or b) it's something sufficiently rare or minor to wait for 10.4.2. I'd probably bet on it being the second one.
A few Safari bugs, yes... also there's apparently something about syncing Address Book data with
So, as you said, it's "right on schedule." I daresay this update being released just now is mainly about getting over the psychological barrier some Mac users tend to have about upgrading to a "10.x.0" release, and also to tidy up some loose ends for Final Cut Studio (which, if memory serves, should ship soon and originally listed 10.4.1 as a requirement until DVD Player 4.6 came out separately). I realize they can't fix everything in one go, but it could be that this release came just a bit early to meet the deadline imposed by Final Cut Studio.
-Frank
I couldn't agree more... I'd love to see lossless music on iTMS too. But I'm fairly certain they'll never do it because of the old burn-and-re-rip DRM hole.
As it stands now, if you burn your 128k AAC purchase from iTMS to a CD and re-rip the result to strip off the FairPlay DRM in an Apple-sanctioned manner, you've either a) lost some quality along the way by using another lossy format to re-encode, or b) grossly oversized the file by using Apple Lossless to re-encode the previously lossy material. But if they offer lossless tunes for download, then that same process will result in a perfect DRM-less copy (unless of course faulty hardware or something similar caused a bit or two to get lost along the way).
They'd never be able to sell the record companies on that one. So, unfortunately, I doubt we'll ever see lossless downloads from iTMS unless they prevented them from ever being burned to CD (in which case a lot of the desirability flies right out the window).
-Frank
PCI-E isn't really necessary - you'd be hard-pressed to fully use the available bandwidth on graphics cards - AGP 8x is damn fast.
I directly addressed this in my post, right after the part where you conveniently stopped quoting me. Yes, AGP 8x is "damn fast." No, current graphics chipsets don't really need all of its bandwidth, let alone that which would be provided by PCI-E. But as I said originally, it's "a performance gap (real or perceived, it matters not in the marketplace)." Even disregarding that, the sad fact is that PCs far outnumber Macs in the world, and their transition to PCI-E is well underway. Once that's done, graphics card makers like ATI or nVidia will be even more reluctant to sink money into Mac graphics card development. It's a small market now, and in the future it will be a small market that is even more expensive to develop for because of the need to tie new graphics chips to the older AGP 8x bus that PCs no longer use. Graphics cards for Mac could become even more expensive than they are now, the delay between introductions of the latest chips for PCs and Macs could become even longer than it is now, and neither of those are good signs for a machine that is meant to "likely last 5+ years" as you said. A machine that is meant to last a long time needs to be designed with an eye toward future expandability, not just current needs.
Seriously, what would you add to make the G5 more enticing? it's arguably faster than any dell available, and will likely last 5+ years, so what's missing?
Enticing to new buyers? Nothing. It's great. Enticing to current owners of Macs, or to power users of PCs looking to switch? Only a few things, but I already named them in some detail in my original post.
-Frank
We refresh our product lines roughly once every nine months. We've been doing it that way for years now. Why is this a surprise?
Not to be rude, but I don't recall ever saying it was a surprise...
Hopefully nothing. "Power Macintosh G5" is a brand item for us. We don't want to release a product and then suddenly drop it. Instead, we want to release a product and maintain it for several years
In the computer industry, "maintaining" a product for several years sometimes means updating components to be more current technology. How would offering faster RAM, for instance, damage the brand recognition of "Power Macintosh G5"? If anything, such a change would strengthen the idea that the PowerMac G5 is truly a world-class workstation computer and provides the best performance available. Saying that "hopefully nothing" changed in the G5 in nearly two years makes it sound like Apple wants a computer to be like a piece of fine furniture. Design it once, change it never. But that's not a very realistic way for the power user market the G5 is apparently targeted towards to view a machine.
Bottom line: We don't just roll out whole new products willy-nilly. Part of what we sell our customers is stability. One of the things you know when you buy a Mac -- most of the time -- is that the thing you buy isn't going to be just totally lame next month.
Few people would call a logic board revision a "whole new product." Saying that the suggestions I made regarding PCI-Express or DDR2 would be rolling out a whole new product is like saying that the Sawtooth G4 was a "whole new product" because it had AGP instead of the PCI graphics like the prior Yikes! model had. Or that the G4's later move from PC133 SDRAM to DDR SDRAM was a whole new product. They were still called G4s, and they still sold well. I don't suppose the brand recognition of a "Power Macintosh G4" suffered for it. And in any event, because of the predictable 9-month upgrade cycles you mentioned, only a few unlucky buyers on the edge of an upgrade cycle would be faced with having a "totally lame" machine soon after their purchase.
My "Bottom line": Apple can produce and has produced more significant performance upgrades to their pro line than either of the G5 upgrades thus far has been. The real issue is that the G5 seems to be suffering from a premature onset of the sluggish development pace that characterized the G4 late in its run. And that's a worrying sign to Mac users like myself.
-Frank
Actually, one thing I think is even more damaging to Apple than the "rumor mill" is the perceived stagnation of the PowerMac lineup. To an outsider (like a consumer in the market for a machine), what has changed in the G5 in nearly two years since its introduction?
1.CPU and FSB speeds have increased.
2.Bigger hard drives.
3.Faster Superdrives.
4.Faster video cards.
I've left out some details obviously, but those are a few major features someone might look at when considering a G5. Of the things I've listed, only the first item (faster CPUs) can't be readily added to either of the two previous generations of G5 at a very nominal cost and with little effort. So while the new PowerMacs are quite impressive machines and are certainly great for first-time buyers, what they may fail to do is to entice any current G5 owners to upgrade to a newer model. And considering that the first G5s appeared nearly two years ago now, that current-owner-looking-to-upgrade-soon market is going to be sizeable in the relatively near future.
What would entice one of those owners to buy a new G5? Perhaps moving to PCI-Express graphics. Perhaps moving to DDR2 RAM. Adopting any of those technologies that 1) the PC market already has, and thus yields a performance gap (real or perceived, it matters not in the marketplace), and 2) that can't be retrofitted onto an older G5 logic board for low cost and effort. Lacking dual-core CPUs wouldn't even be very disappointing if the rest of the architecture surrounding single-core chips got a bit of a revamp.
In short, *that* is the sort of upgrade that will capture both new and existing Mac users to buy a new G5, and the lack of such an upgrade is probably as big of an influence as the "rumor mill" you blamed in damaging Apple and keeping PowerMac sales a bit low. As a current G5 user, I for one hope such an update comes soon. My checkbook will be ready.
-Frank
Actually, following the House bill's description on thomas.loc.gov leads to a page, showing that Sen. Hatch is the main sponsor of the Senate versions of this legislation.
I would guess that as Sen. Hatch is the one who originally introduced the bill to the Senate, he would vote in its favor when the time comes... assuming of course that he's not too offended by any changes made in the House.
-Frank
I might mention that Kevin Gerich's widget set makes Firefox's HTML controls look much more presentable on Mac, in my opinion. It's not quite the same as having native Aqua widgets, but it's a start. Granted they aren't bundled with the application by default, nor do they solve any of the other OS integration issues you mentioned.
That having been said, I agree with the assessment that Firefox for Mac has a lot of catch-up to do to match Safari in terms of aesthetics. It's one of the biggest cons of choosing Firefox on the Mac platform. Safari, as Apple's own in-house effort, gets a level of fit-and-finish with the rest of the OS that third-party developers can have a tough time matching.
On the other hand, the biggest pro for Firefox on Mac (in my opinion) is the expandability. Safari doesn't have Adblock, BugMeNot, or any of my other favorite extensions. Even Camino doesn't support them. So in my case, I choose expandability over aesthetics and use Firefox as my default browser on Mac.
Ideally though, it would be possible to have both. Maybe in time and with further Firefox development.
-Frank