Solipsism Gradient

Whew. Apple (formerly “Computer”) Inc. has managed to surprise almost everyone. Yesterday I sat up late reading comments and analyses, and there’s a bewildering array of information (and FUD) out there – and we’ve seen only the tip of the iceberg yet.

First of all, there was no mention of the options issue, although Steve Jobs touched on it briefly in a TV interview afterwards, saying he was confident that it’s not an issue anymore. Investors seem to agree; AAPL stock soared over 7% during the day and gained a little more after hours. As the owner of a modest few shares, I’m very happy. Also, Jobs looked healthy and fit, and no mention was made of health issues or resigning/retiring, and we’ll hopefully hear no more of that. I guess he looked haggard lately because of all that secret twiddling away on the iPhone design…

My personal predictions didn’t fare well on the whole. Of course, like everybody I thought it’d be a Macworld (note lowercase “w”, sorry, must be some ReFlex to write “W” there) keynote, but it turned out to be an Appleworld keynote. So, no Mac news, no Leopard news, no iApps news, no .mac news. Except indirectly, which is of course also significant. Many people were disappointed with this, of course, and if I wanted to be superficial I could say so too; after all, here are two devices I’d never use myself (a phone and a weird TV interface thing), and my pet subjects weren’t even mentioned. A third device, the revamped Airport Extreme, didn’t rate a keynote mention; neither did the 24% price drop on the Airport Express; the general migration to 802.11n was glossed over very briefly.

On the software development front, there’s nothing new either; the only exception is a new document called “Introducing Dashcode“. No new Leopard seed.

So why all the uproar over a couple of consumer electronics items which should have been introduced at the neighboring circus, the CES? And where’s the Mac news?

I’m declaring myself disqualified to write much about the i, oops, AppleTV. As I’ve said before, I don’t watch TV much, and of course all the stuff like the iTunes store, or even PVRs, aren’t available here – and they’d probably sink fast if they depended on me as a customer – I don’t need or want to be “entertained” the whole day, or even at all when I’m home. To me the AppleTV looks like a huge non-portable iPod without a screen, and that’s it.

The iPhone looks a little more interesting, and I might even look into buying a second or third generation version of it. If it truly does have GPS, as some say (others say it uses less precise cellphone-based localization techniques), it will be very useful for trips. Even so, it remains to be seen how the service side of that will work out. Currently the iPhone is tied to a single service (Cingular), it won’t be out until middle of the year and the operating costs haven’t been disclosed, so it’s pretty much a limited-application, US-centric device, and will remain so until 2008 or even 2009. So I won’t talk much about the phone aspect either.

But when you look at the whole picture, it becomes much more interesting. If it were just a cellphone with a built-in iPod (or vice-versa), even Steve Jobs would have been hard put to spend nearly two hours demonstrating it. On the contrary, we really saw just the highlights, there was very little hard technical information, and the implications are far-reaching. Let me touch on some of those in no particular order.

The iPhone runs “OS X” (note the missing “Mac” in there). Yes, it’s an embedded version of Mac OS X, and it’s stripped-down to fit on the device. All the major stuff seems to be in there – generic application support (at least as far as Cocoa, WebKit and networking are concerned), high-end graphics support or a works-well-enough fake thereof, and of course a cutting-edge user interface. How much did they strip out? Details will have to wait until June, I’d say. Not even the exact processor type is known; people tell me an Intel processor was mentioned during the keynote, but Intel makes a huge range of embedded CPUs, ranging from legacy 80186s to brand-new Core 2 Duos. There’s also a huge range of supporting chipsets available. I suppose it uses an integrated GPU with some OpenGL support, but it needn’t be too powerful, since we’re talking about a comparatively tiny 320×480 screen here. I also wouldn’t be surprised if it used some as-yet-unreleased Intel chip.

Still on the hardware, the iPhone will come in 4GB and 8GB varieties. Of course they’re talking about the flash memory capacity here (or is it a tiny hard drive?), and no mention is made of built-in RAM and boot ROM. I suppose RAM needn’t be too large – 256MB sound good to me. For comparison, I believe the iPods have 32MB. The OS and the applications would be stored, probably in compressed form, along with the songs/pictures/videos, but it wouldn’t use more than 10% of that. As with the CPU, Apple has been careful not to talk about exact capacities.

So will Leopard be officially released as “OS X 10.5”? Well possible. It seems safe to say that the iPhone OS is some sort of “Leopard Lite”. The user interface looks like the Finder has been stripped out and Dashboard put in its place, with every widget running maximized (to coin a term ). In fact, Steve Jobs showed that widgets are supported, which makes sense. Widgets run off JavaScript and WebKit, with lots of underlying Cocoa help, so they’re relatively lightweight. Even Safari and Mail could be turned into widgets once you put more direct support for their UI into Cocoa, so I presume that’s what Apple did. So is the recent release of the Dashcode beta a coincidence? Time will tell… for now, no word on whether Apple will allow random developers to do iPhone apps, or whether there’ll be an invitation-only selection process like they’re doing with iPod games. This year’s WWDC promises to be even huger than the last two, mark my words; I certainly intend to be there. Estimates of Leopard’s actual release date vary wildly from February to June, although April/May was my personal bet before the keynote; now, my feeling is Leopard and the iPhone will be released on the same day. Perhaps around the end of May; delivering early is part of the new Apple not-only-computer philosophy.

You’d think the rumors of a Tablet Mac would have died by now, but people are still calling for one. But the iPhone is sort of a Tablet Mac; I can well envision a second or third-generation device with a 9″ or 10″ screen, and a more mature gestural interface. This would need more advanced screen and battery technology than what we have on the market now, but next year it’ll be another story. Would Apple just turn the screen around on a MacBook and require users to use a stylus, or keep the current UI which is designed for a mouse+keyboard interface? Of course not; that’s why current Windows-based tablets remain a niche product. The new multi-touch UI looks just like what the doctor ordered.

My feeling is that the whole gestural interface, multi-touch screen, animation-centered philosophy will percolate back into mainstream OS X/Leopard and that what we’ve actually seen was a preview of one UI mode in Leopard; perhaps the Simple Finder/kiddie mode, at the very least the Dashboard interface. Why did the Finder seen at the WWDC Leopard preview look just like the ho-hum old one? Perhaps it’s destined to be an optional install for traditionalists. At this point I think we should try to extrapolate what the iPhone interface would look like on a large screen. Imagine the entire palm rest on a MacBook turned into a multi-touch interface. Optional keyboard? I doubt even Apple could pull that off, but it’s not impossible.

Looking back at Apple’s recent patent filings much of what we’ve seen was already revealed, especially the multi-touch part. Is the iPhone really made of that radio-transparent ceramic, and is that the secret of the supposedly smudge-free surface? If so, a larger tablet version is just a matter of time. One item which stayed in the labs was the screen-as-camera patent; that might be the reason the iPhone doesn’t (yet) have a second camera in front for video chatting.

No new iPods were announced for a good reason: the iPhone now is the high-end iPod. Look for future iPods, except for the screenless ones like the shuffle, to become iPhone Lites; at first in styling, later in hardware platform.

Regarding market share, Apple is looking for 1% of the global cellphone market next year, which seems to mean 10 million devices sold. However, as with the original iPod, that’s not the right point of view. The iPhone competes more with the current “smart” phones, of which 6.5 million were sold last year. Even supposing those figures don’t change, and nobody migrates to the iPhone, it still seems reasonable to assume Apple will capture half or more of that market. I suppose the competitors are already tearing their hair out at this point. Read Bill Gates’ CES keynote transcript just a day before for some chuckles:

…mobile phones. This is an area where we’ve made tremendous progress. This year we have some of the hottest selling phones in the marketplace, and the cool thing for me about those phones is it’s not just about phone calls, although we do that great, it’s not even just about e-mail, since that was the next round of things people wanted to be able to do, but it’s also about IMs, it’s about movies, it’s about TV, it’s about music, it’s about connected entertainment on my phone.

So the features we’re delivering in these exciting products are bringing that to market. If you think about Cingular’s Blackjack from Samsung, Cingular’s Treo from Palm, the T-Mobile Dash from HTC, and Verizon’s Motorola Q, those four alone are leading, cutting edge designs that are driving tremendous market share advances for Windows Mobile…

Talk about instant obsolescence…

I known there’s much more to talk about, but this post is already too long. The upside is, we’re looking at lots of “special events” over the next months, each (hopefullly) with its own little stock boost.

MacWorld predictions (and the parallel tracking of the options “scandal”) are growing ever more frenzied, and even starting to overlap, with a pronounced “Chinese whispers” effect, even stronger than usual. I’m not a lawyer or stock/options expert, but my bogometer just about pegged. Frankly I can’t see how Steve Jobs resigning (or even taking some time off for “health reasons”) would help Apple stockholders in any way. FWIW, I think nothing serious will come of this… and I’m not selling off my AAPL stock before the keynote.

Looking over the spate of recent rumors, it struck me that it’s very likely that a future small Apple device would feature an “embedded” Mac OS X. After all, embedded CPUs are constantly becoming more powerful, and RAM and flash memory capacities on embedded devices are also growing – and Unix-like OSes are known for being well-suited to embedding.

At first glance, Mac OS X is a huge beast. You need several gigabytes drive space for a normal install, and half a gigabyte of RAM is now standard on Macs – meaning, double that is the minimum for serious use. However, most of that stuff is necessary for general use. If you look inside the system folders you’ll see tons of drivers for all currently supported hardware configurations, tons of frameworks for all sorts of applications, and lots of extra apps and utilities.

In contrast, any embedded device running Mac OS X would have a stripped-down version of the kernel (most probably supporting only one CPU), the exact selection of drivers and frameworks needed for that device, and one single application controlling the user interface. Supporting additional applications would be relatively easy once their focus is defined. Looking at it in another way, this describes the original Mac very well… and it had 64K of ROM, 128K of RAM and 400K of disk space. That’s much less than the iPod has, of course.

So how small could Mac OS X be squeezed in order to fit into the iTV, or into any rumored handheld Apple thingy/paradigm smasher? That’s hard to say. Offhand I’d estimate 64 to 256MB of flash memory, anything from half to double that of RAM, and that without raising the price too much, given the volume discount Apple could get.

Scott Stevenson just posted “The Year in Mac Development“, an excellent summary. Two comments on specific points…

Re: interface guidelines:

If you need one rule to follow, make it this: don’t introduce new behaviors for existing controls. Don’t make a checkbox act like a push button or a slider act like a scroll bar. That will confuse users. If you need new behavior, make a new control. And when in doubt about UI standards, just do as Apple does.

I’d just sent a build of XRay II for a developer friend to look at – a very early pre-alpha, I hasten to add – and he called my attention to a detail on the UI. While replying to this, it occurred to me that almost all of my UI was custom controls; the only exception were scrollers and text fields. However, my friend hadn’t caught on to this at all, as the custom controls looked and behaved for the most part like Apple’s controls – or at least, like users think Apple’s controls should behave. A good example for this is of course RBSplitView; you can twiddle it to look just like any variation of Apple’s split views look, while underneath it’s all new code.

So, I’d amend Scott’s comment here to “don’t introduce new behaviors to existing controls unless they’re expected. Make a checkbox act as a better checkbox, not like a pushbutton.” Flashy visuals may well be nice or even appropriate in specific situations, but they’re liable to get in the way of the “just works” experience.

Re: marketing shareware controversies:

Perhaps most importantly, these events took the idea of buying independent Mac software out of the shadows and brought it into the community’s consciouness. A lot of the stigma of supporting smaller developers has been lifted, hopefully improving things for everyone involved.

For various reasons, I’ve stayed away from (or wasn’t invited, which was just as well) these new marketing events, tempting though they looked at first sight. They seem directed at a more general public than what is intended for my main application. Still, it’s interesting to note that my downloads and sales have picked up noticeably in the last 6 months. No idea if this is due to the effect Scott comments on, though…

Coming back to MacWorld, here are some more expectations I saw on various sites:

– iApps. Well, of course they’ll be improved as usual, and the Leopard versions will use stuff from the new frameworks. I still can’t understand why so many people turn this concept completely around and mention small app changes when they talk about the OS releases, though.

– iPods. Another area where incremental improvements go without saying. Apple might have a reply to the Zune ready, perhaps with a larger screen for video, perhaps with the patented touchscreen/body. Still, I’m not excited about this.

– Some people are – maybe only half in jest – suggesting that Steve Jobs will announce a career change. Perhaps a Bill Gates-like lateral promotion to “Chief Technologist”? There might something to that. The recent options flap showed that Wall Street’s expectations of a CEO’s functions are increasingly dissociated from what Jobs really does (or should be doing, anyway) at Apple. Of course you and me know that Tim Cook, the COO, already does the dull back-end operations stuff, but do the analysts understand that? I don’t think so.

Posted by gracion:

Rainer Brockerhoff wrote:

– .mac is dead. It’s never worked well (or at all) outside the US, as far as I know. It’s an expensive embarrassment. Apple could close it outright, sell it to Google, or allow people to operate their own sync servers.

Happy New Year, Rainer! Very good summing up of the year, which I agree with, except I’m not sure I want to kill .Mac. Here’s why:

It has a lot to offer as a seamless extension of Mac OS X, especially for nontechnical users. (1) Backup works, is flexible but dead-simple to use. (Yes Time Machine will change all that, sort of). (2) iDisk, ditto. (3) Photo web publishing from iPhoto is again seamless. (3) OS-integrated syncing of a variety of things (room for improvement though) (4) an interesting, already-working API for iDisk access, store-and-forward, and syncing. (5) Amazon discount (6) Comcast, Charter, and other ISPs are starting to blacklist my small independent ISPs mail server willy nilly. .Mac email is my way around these.

One of my very nontechnical Mac user friends had all sorts of trouble doing things until I got them on .Mac. I suspect that is true for a lot of Mac users. Admittedly, I haven’t checked out alternatives. Is there a “GoogleDisk” item I can add to the Go menu in the Finder?

Anyway, I’m looking forward to fun at Macworld, iPhone or noPhone!

Enjoy!

Heh. Seems the year’s over already and I almost didn’t notice.

Looking back, it’s been a busy and surprising year. I traveled more than I’d planned and wrote less. XRay II saw a lot of progress in fits and starts, but my plans to release a public beta this year didn’t work out – mostly because I had underestimated the back-end work necessary to actually save data.

On the Apple front, the year has been busy. No iPhone from Apple. Zune out, Vista out (sort of), AAPL options scandal, Leopard, Mac Pro, the Intel migration has been completed, lots of security flaps, new laptops… it’s a long list, so long I don’t feel like finding all those old links. Looking back, what surprises me most is that Apple doesn’t seem to be as interested in virtualization as I felt they should be.

And of course MacWorld is just a little over a week away. Rumors are already flying fast and furious, of course. Here are some things I believe to be more likely (not that I have any inside knowledge, I hasten to add):

• New Apple displays, with built-in iSights and microphones.
• New Mac Pro with 8 cores, probably with a new case design.
• The transition is over, and people are now sure the Intel Macs are “really Macs”, so new case designs are overdue across the whole line, although in the case of the laptops I’d say that’ll be really hard. Thinner and better/larger screens of course, but there’s only so much you can do with minimalism.
• Leopard? Perhaps we’ll finally see some UI changes. My tip for the release date is March/April.
• iTV, no idea in which direction they’ll take that; I watch very little TV. Regarding the name(s), it would make sense to go away from the whole iThing.
• whateverPhone: I don’t use a cellphone, so the basic idea leaves me cold. Unless Apple breaks the entire paradigm with some sort of VoIP breakthrough, it’s bound to be some sort of weak US-only experiment. Let’s hope they don’t do that. I also see no sense in having music capabilities built-in as a default. Opening it up to developers in a big way would be excellent, and the recent rumors of a stripped-down, embedded Mac OS X dovetail with that.
• .mac is dead. It’s never worked well (or at all) outside the US, as far as I know. It’s an expensive embarrassment. Apple could close it outright, sell it to Google, or allow people to operate their own sync servers.

A few years ago, when I was still getting the idea of this blogging thing, I made a serious effort to stay “in the know”. I read most of the Mac websites, I had 390+ feeds in my news reader, I posted links to interesting blogs, I tried to comment on the hot issues of the day. No idea how successful that was (depending on your definition of “successful”), but one side-effect became apparent after a year or so: no useful work on my applications got done. There are just so many nanoseconds in a day.

Perhaps the main cause of that was my overly-zealous polishing of each sentence – writing in what is, after all, my fourth language isn’t that easy – but if there’s any obsessive-compulsive polishing that must be done it would be better applied to my code than to my text. Right? On the other hand, there are people who tell me they like reading what I post here, if only to keep up-to-date with my trips. And the whole thing was, after all, just a sideshow to my support forums… no sense in closing it down.

So, I’ll probably not comment after the fact on most of the various issues du jour here… there have been an awful lot of them lately. I won’t even take the trouble to find links to them now. Let’s see, there was the AirPort security thing, the HIG-is-dead/Disco thing, the MacHeist controversy, the iPhone came-out-but-not-really flap, the options scandal is still going on, I still can’t comment on Leopard, yadda yadda.

My late father worked at a large company and he used to tell with some relish a story about how he used to sort the requests that crossed his desk into “not urgent”, “normal” and “extremely urgent” piles. His usual mode of operation was to ignore the “extremely urgent” stuff until someone asked after a particular item at least twice; it turned out that most of them were never followed up at all! The lesson has served me well. Many of those hot issues have a short half-life, emitting lots of sparks but decaying very soon into plain, dull lead. Nothing like letting a few weeks or months pass to get the proper perspective…

In the meantime, yes, suddenly I’ve been able to get lots of polishing done on XRay II. Keep tuned.

I’ve written about Apple’s use of the TPM chip before. My basic conclusion was, there’s no evidence Apple is using the chip for anything sinister, or at all in current versions (Tiger). However, I also said Apple should use the chip as a basis for secure vitualization in Leopard:

…Apple should write a fully trusted hypervisor into the EFI (using the TPM) and run everything inside virtual machines, including Mac OS X for Intel itself. Booting some version of Windows into a second VM would be easy, then, and there wouldn’t be a full version of Mac OS X for Intel for people to run on standard PCs either. I don’t think dual-booting is a good solution, I believe Apple was just testing the waters with BootCamp.

I still think virtualization is a good idea… however, there’s new evidence that Apple doesn’t think so, or at least not in conjunction with the TPM chip.

First, ifixit posted a disassembly of the new Core 2 Duo MacBook Pros, with zoomed-in photos of the logic board. They’re not detailed enough to show all IC part numbers, but I can say with some confidence that there’s no TPM chip at all. However, to the right of the RAM socket in the second picture, there’s an empty space for a 28-pin flat-pack IC – just the size of the Infineon SLB9635TT chip found on all previous Intel Macs. I’ve been searching for a similarly detailed picture of the Mac Pro’s motherboard, with no luck so far.

Second, Amit Singh of Mac OS X Internals fame – which I bought and read recently, BTW – has posted, in his usual precise style, details on how to use those Macs’ TPM chip. Here are some salient points:

The media has been discussing “Apple’s use of TPM” for a long time now. There have been numerous reports of system attackers bypassing “Apple’s TPM protection” and finding “Apple’s TPM keys.” Nevertheless, it is important to note that Apple does not use the TPM. If you have a TPM-equipped Macintosh computer, you can use the TPM for its intended purpose, with no side effect on the normal working of Mac OS X.

…

At the time of this writing (October 2006), the newest Apple computer models, such as the MacPro and the revised MacBook Pro do not contain an onboard TPM. Theoretically, Apple could bring the TPM back, perhaps, if there were enough interest (after all, it is increasingly common to find TPMs in current notebook computers), but that’s another story.

He then goes on with very detailed instructions on how to write, install and use a device driver for the TPM chip.

All this is very interesting, but as the TPM isn’t anymore standard equipment you could rely on finding on any Intel Mac, this is more an academic exercise. I doubt that Apple will implement anything important in Leopard that won’t run on the new Pro machines, so no trusted hypervisor for me. Ah well…

Adam has posted a response on the installer authorization issue:

I don’t believe they’re even calling that function to gain root, honestly, because it follows the authorization file. It can’t not. They’re doing something else and I believe that’s a red herring here. There’s no way to call that function and have it not consult the database, so they’re doing something internal to get around it. Be that a SUID program somewhere or some private call, they’re getting around the clause in authorization that says the user needs a password.

Well, that’d be a surprise to me, but it’s not impossible. I’ll try to find some time to do a test package and some rooting (oops!) around inside the Installer before writing more about this.