• Explore the magic and the mystery!



  • Newsletter Issue #505

    August 2nd, 2009

    THIS WEEK’S TECH NIGHT OWL LIVE RADIO UPDATE

    So we’re just a few months away from the arrival of Windows 7, but does anyone rarely care? You read extremely positive reviews about how much faster it actually is, but the one set of benchmarks I read, on CNET, a known Microsoft supporter, showed very little measurable difference when compared with Vista. The one significant advantage was in shut down speed, which simply means you can get rid of its trashy interface faster. That’s the ticket!

    Staying on topic, on this week’s episode of The Tech Night Owl LIVE, we had a return visit from cutting-edge commentator Daniel Eran Dilger, of Roughly Drafted Magazine. He talked about the dimming prospects for Windows 7, netbooks, the possibilities for an Apple-based tablet computer and lots more.

    We then entered “The David Biedny Zone,” where our Special Correspondent holds forth on a variety of topics in his inimitable fashion. First on the agenda was the fate of Moore’s Law, and what chip makers had to do when increasing gigahertz ratings ran out of steam. David also discussed Microsoft’s failing prospects, the possibilities for an iTablet computer from Apple, the iPhone, and he also delivered a preliminary analysis of Apple’s new Logic Studio recording studio software.

    This week on our other radio show, The Paracast, we remember noted UFO researcher Richard H. Hall, who died recently. Featured guests include some of the esteemed investigators in the UFO field who knew him well, including Jerome Clark, Don Ecker, Paul Kimball, Dr. Bruce Maccabee and Kevin D. Randle.

    Coming August 9: Learn about the amazing life story of writer, producer and modern mystic Walter Starcke.

    Coming August 16: Paranormal talk show hosts, investigators and experiencers Paul and Ben Eno discuss their ongoing research.

    Now Shipping! The Official Paracast T-Shirt. We’re taking orders direct from our new Official Paracast Store, where you can place your order and pay with a major credit card or PayPal. The shirts come in white, 100% cotton, and feature The Paracast logo on the front. The rear emblem states: “Separating Signal From Noise.” You can get them for $14.95, each, plus shipping, and you can select from most popular sizes.

    THE END OF POWERPC SUPPORT IS CLOSE AT HAND

    I have to think that the PowerPC never really attained its true potential, but that’s not Apple’s fault. First Motorola and then IBM failed to deliver the speedier chips Apple needed to keep up with the rapidly-advancing X86 platform. Both AMD and Intel kept moving ahead, whereas Apple’s chip partners languished and failed to deliver the right chips at the right time.

    Take the original announcement of the G5. Steve Jobs introduced an on-screen presentation featuring IBM’s state-of-the-art processor fabrication plant, and how they had developed technology that would blow the competitors from Intel away. He even had the temerity to promise that, a year hence, there would be a 3GHz version. Outstanding!

    But the G5 had some serious deficiencies, most notable of which is that they ran real hot. So the most powerful Power Mac G5 required multiple fans and liquid cooling to keep the box at a safe operating temperature. Some folks actually suffered coolant leaks, which meant that their Mac towers were “toast,” as far as repairs were concerned.

    Unfortunately, IBM couldn’t tame the G5 sufficiently to reduce the massive cooling requirements, which pretty much put a serious damper on the prospects that there’d ever be a note-book version. So what was Apple to do?

    Although it had long been rumored, Jobs didn’t reveal until is keynote at the 2005 WWDC that there was, in fact, a secret Apple project in place to develop a version of Mac OS X that would run on x86 processors from AMD and Intel. Although the essential details had already been published by the mainstream media ahead of the official announcement, it was now official. Apple had made a deal with Intel to move the entire Mac line to their processors by 2007.

    It’s really smart to promise less and deliver more. Apple does that routinely with their conservative financial guidance for future quarters. But with the Intel transition there was the danger that Mac sales would stall as customers waited for the new chips. Beginning in January 2006 and ending in August of that year, the switchover was handled in record time, way ahead of even the most optimistic estimates.

    Sure, there were some growing pains. The early Intel-based Mac note-books tended to run hot, though subsequent updates to improve cooling fan efficiency eradicated some of the worst of the ills. Subsequent models took advantage of improvements in Intel’s processors, which routinely operated at even lower temperatures. My early 2008 17-inch MacBook Pro never seems to get more than warm to the touch, even here in the blazing hot southwest.

    Performance proved to be a revelation, and as Intel added cores and greater processing efficiencies, it only got better. Finally, Apple had a processor roadmap they could depend upon.

    At first, compatibility with PowerPC software was enabled with Rosetta, essentially an emulation tool, but new applications were Universal, which meant they would run on both the PowerPC and Intel. However, you could see the handwriting on the wall as more and more apps became Intel-only.

    Consider the return of Adobe Premiere Pro, which departed the Mac platform after Final Cut Pro gained ascendency in the film and TV industries, and also dominated sales. That and a few other Adobe apps would only run on x86 computers, either Mac or Windows. Some games were readily ported to Mac OS X, but not the PowerPC version.

    In 2008, when Apple first announced the arrival of Snow Leopard, a major enhancement of Leopard, they didn’t actually specify system requirements. But a few developers leaked the news to the Mac rumor sites that Apple’s prereleases eschewed PowerPC support. The official confirmation came at the 2009 WWDC, when Apple released actual pricing and the final feature-set, along with the fact that, yes, it would only run on Intel-based Mac hardware.

    The nail in the coffin for the PowerPC has become clearly visible in recent months. More and more apps don’t support the older processor family. Apple’s newest professional content creation app suites — Final Cut Studio and Logic Studio — are also strictly designed for Intel chips. It’s only a matter of time for the consumer apps to follow.

    However, there’s no reason to feel betrayed or upset by these developments. It is over three years — an eternity in the computer business — since Apple built its last PowerPC Mac, the G5 family, and thus it makes perfect sense that they want to concentrate on the future and not prolong past glories.

    For those who might feel abandoned, don’t forget that a PowerBook G4 is dog slow compared to even the entry-level MacBook, which retails for $999. Even the $599 Mac mini can deliver performance that exceeds the most powerful Power Mac G5, except, no doubt, for hard drive performance. What this means is that the price of admission for an upgrade to allow you to run current Apple software isn’t terribly high.

    I suppose it would have been nice to see Apple continue to support the PowerPC for a year or two longer. However, that’s not a cheap thing to do, and providing products for two incompatible processor families means there are apt to be inefficiencies along the way, including the extra disk space taken up by Universal code.

    In the end, I don’t miss the PowerPC all that much. It’s heyday was of relatively short duration. For a while, it even beat the Pentium in those knockdown drag out bake-offs that Apple would stage at press events. But when Apple’s chip partners gave their business lower and lower priority, you could see that Apple really had little choice but to go along with the rest of the PC industry and embrace x86. If you look at the profits and the sales figures, though, it really turned out really well, don’t you think?

    SO IS ALEXANDER GRAHAM BELL REALLY SPINNING IN HIS GRAVE?

    I have a few close friends who will probably never buy a cell phone. That might seem awfully retro as more and more young people dump landlines. But I can see the point. You see, a traditional landline phone, connected to a standard packet switching network, sounds pretty decent. Although of a limited frequency range, voice quality is robust and crystal clear.

    The various Internet phone (VoIP) providers come awfully close, and the technology is getting far more reliable as well. But when it comes to a strictly wireless connection, voice quality seems to be the last priority for most of the carriers in the U.S. I can’t, of course, speak to the situation in other countries.

    Even on the supposedly most robust network in this country, owned by Verizon Wireless, the best connections are frequently overrun by digital haze. Voice quality seems compressed and trashed until it becomes barely understandable. Even when disconnects are few and far between, getting audio that matches that of two tin cans connected by a single stretched wire remains an unfulfilled dream.

    I sometimes wonder why more people don’t complain, but I realize that we all have our priorities. Besides, being able to understand a caller on a wireless phone ought to be sufficient. Kids really use them mostly for texting anyway, so sound quality gets shuffled to the bottom of their list of priorities.

    I have had service with the top three — AT&T, Sprint and Verizon Wireless. Although voice fidelity varies somewhat among these carriers and from handset to handset, the differences are not significant. I suppose we should be grateful that dropped calls aren’t frequent and be done with it.

    Certainly AT&T gets its lumps from iPhone users and deservedly so. They are late to the party with multimedia texting, or MMS, and what about being able to tether your iPhone to a computer, which means using it as a broadband modem? Worse, their 3G coverage in major cities still leaves something to be desired.

    But when you read all of the complaints, you seldom see anything about making voice quality as good as a standard phone. Shouldn’t that be a priority? You are certainly paying a lot more money to dispense with the wires, and the hardware — even if it’s offered free of charge — is a whole lot more costly on the long haul. So why shouldn’t you be able to always hear the person at the other end of the line, and be assured voice quality is state of the art?

    But I’m no doubt barking up the wrong tree. My priorities certainly don’t resemble that of most other people. Wherever he is now, I wonder what Alexander Graham Bell would think about how his invention turned out.

    THE FINAL WORD

    The Tech Night Owl Newsletter is a weekly information service of Making The Impossible, Inc.

    Publisher/Editor: Gene Steinberg
    Managing Editor: Grayson Steinberg
    Marketing and Public Relations: Barbara Kaplan
    Worldwide Licensing and Marketing: Sharon Jarvis



    Share
    | Print This Issue Print This Issue

    5 Responses to “Newsletter Issue #505”

    1. dfs says:

      So far, Apple’s move over to Intel chips has seemed like a smart move and it was nothing short of brilliant how smoothly they managed the transition. It’s equally brilliant that we can now run OSX and Windows from the same box, thanks to this decision. And three years’ further support for the G-5 is quite adequate, if Apple wants to pull the plug that’s fine by me. But there’s still a huge question hanging over the whole Apple-Intel partnership. Multicore processing is in essence an Intel vision (embracing both the Mac and PC platforms), a way of getting around the fact that it’s increasingly difficult to speed up processors in any other way. They’ve convinced Apple to buy into this vision, as will be seen in Snow Leopard. But Intel also needs to get the software industry to get aboard the train if it’s ever going to leave the station. Within a couple of months, I’m going to have Snow Leopard running on my four-core Mac Pro and I’m going to feel like I’m all dressed up with nowhere very much to go, until when and if the big software companies like Microsoft and Adobe bring out multicore-friendly versions of their products. How long is this situation going to last? For just a few months? Permanently? I have no idea how they are reacting to Intel’s multicore vision. Does anybody? So the ultimate success of the Apple-Intel partnership really rests in their hands. This is a situation, where Apple’s future success depends on decisions made by other corporations, that makes me more than a little nervous.

    2. Ilgaz says:

      Gene, the current Mac Mini (with nvidia gfx) has amazing hard disk performance. I am absolutely shocked since the HD coming with the Japanese version is not so successful Fujitsu HD.

      I have seen numbers like 90-100 MB/sec in certain occasions. I didn’t run a benchmark yet, all from Activity Monitor. We also have 2 G4 Mac Minis. You can actually figure the Mot&IBM was going nowhere regarding desktop&mobile, especially mobile when you use one after another.

      I think, Leopard only apps will be continued to be maintained and we may see even more apps in Utility department on PPC but there is no way OpenCL etc. powered apps will be back ported to Leopard/PPC.

      Snow Leopard not being PPC compatible especially regarding G5 (as it is 64bit) has interesting reasons you can trace back to IBM easily. IBM did almost nothing in regards to GPU processing, they didn’t work with Nvidia (in CUDA), the G5 has no performance boost if it is run in pure 64bit (in contrast to x86-64) and in fact, ask the high performance Linux guys, it will even slow down if you needlessly use 64bit binaries all over the place.

      If you compile open source software with hand and watch the compile process, you can see the other reason. For years, Intel and AMD worked closely with developers and gnu to make sure their processor’s features are widely used and automatic optimisation on x86 has reached an amazing point today. Compare it to IBM which still hasn’t freed their XL Compilers for PPC and still trying to sell it, that incompatible thing for 600 dollars. Yes, I speak about OS X version which is unsupported beyond 10.2 or 10.3.

    3. Richard says:

      One might suggest that the performance boost x86 experiences in going from 32 to 64 bit is more a tribute to the chronic awfulness of the 32 bit x86 architecture. The specific problem in this case being a lack of general purpose registers – x86-32 only has eight general purpose registers (by contrast the PowerPC line had 32 such registers).

    4. gopher says:

      It is bittersweet to have to lose PowerPC support. Many still rely on software that needs Mac OS 9, and can’t finance the upgrade of all their software to Intel Macs. Some rely on PowerPC only games and can’t upgrade to Intel Mac versions of the same. Sure there is Sheepshaver, but you still need a PowerPC ROM for that. On the other hand, we are glad to see Apple has pulled out of the recession as one of the few leaders in the computer industry, defying gravity, and making profits left, right and center. At least PowerPC Macs can be self-supported for a long time to come, and Apple usergroups have many who will help:

      http://www.apple.com/usergroups/

      So don’t panic, used and refurbished Macs still have a place.

    5. Ilgaz says:

      @gopher I am sure Leopard and currently offered PPC software for Leopard will be supported for a long time. As Richard mentions, Intel Macs definitely needed a pure 64bit OS with technologies like GPU processing support.

      Look to software from trustable names, down to shareware authors. Always pick them if they are decent enough to support 10.4.11 (as of today). I have always picked such software like Graphic Converter, Vuescan, Opera and I am in safe area now. Of course, Graphic Converter will have a `snow leopard features using` version but Lemke won`t abandon 10.4+ PPC just because he has that version. That is what separates goodly written, supported software from free/trendy stuff.

    Leave Your Comment