Live from Austin!

I’ve been watching reaction to the Apple announcement. On the one side, there are a lot of long-time Mac fans who have been expressing a kind of unfocused pessimism. Something about the move makes them really unhappy at a subconscious level, but they seem unable to express exactly what.

Then on the other hand, there are lots of people saying “Hey, it’s just a computer, it’ll still run OS X and be pretty and be a Mac even if it has a Pentium 4 inside.” I think that’s not quite true, in a fairly subtle way. To understand why, though, you need to understand the Mac religion.

Apple have been forced to make the move because the desktop CPU marketplace, while still competitive, is only competitive around a single instruction set–the one that has gradually evolved from the original 8086. The x86 instruction set dominates the desktop, in CPUs from VIA, Intel, Transmeta, AMD and others. It’s even somewhat strong in the embedded space. Compiler technology for x86 advances far faster than for any other instruction set.

So let’s be clear: x86 is completely dominant. And it blows. That’s what makes the move so hard for many Mac-heads to accept. The Mac has always been about doing what’s technically right, not what’s most popular.

For instance, I vastly prefer the clean simplicity of the 6502 to the ugliness of the Z80, having written code for both. The 680×0 was a joy to write for compared to the 8086 thru 80486.

Moving up out of the realm of processors, SCSI was clearly superior to IDE. USB was obviously the right thing, even if serial ports and ADB were far more popular and USB peripherals were initially almost impossible to find. Firewire is better in every way than USB 2.0 HiSpeed.

In the software layer, the way the Mac filesystem works is a pain in the ass to write for, but the way the system behaves to the end user as a result is clearly the right way. (Programs don’t break when you move them, files launch to the application you last edited them with, and so on.)

In short, the Mac has always been about picking the best technology and doing what’s right. But now suddenly there’s going to be an x86 CPU in the middle of it all–kludge after kludge piled on top of the original 8086 design. And recall, IBM chose that because it sucked, they didn’t want to choose something that might threaten their real computer systems. Worse, the Mac isn’t even going to be using leading-edge AMD 64 bit x86 CPU, it’s going to be a 32 bit Intel processor.

Basically, the Mac community is being served a shit sandwich. It may still be the finest ciabatta bread, the freshest pickles and lettuce–but there’s going to be a huge turd in the middle, and some of us are having a hard time preparing to swallow it, even though we know it may be necessary–because as the saying goes, “Eat shit! Fifty billion flies can’t be wrong!”

Since I know people find my web pages while searching for information about Nikon scanners and Mac OS X, I’d like to offer the following endorsement:

The Ratoc FR1SX Ultra-SCSI to Firewire adaptor works perfectly with Mac OS X 10.3, and doesn’t need any drivers.

Plug the unit in to the back of your SCSI-based Nikon film scanner, and you suddenly have a Firewire-based Nikon film scanner. This can then be used with Ed Hamrick’s excellent VueScan software to fulfil all your scanning needs.

No adaptor drivers, no Nikon drivers, no Nikon software of any kind—so this solution should work fine with 10.4 / Tiger and other forthcoming OS X versions too.

Unlike USB options, the Firewire interface seems to result in scanning being just as fast and reliable as it was with SCSI.

Yesterday was a day of unwanted learning experiences. A less charitable observer would have termed them ‘massive fuck-ups’, but I think I learned plenty.

In particular, I learned that one of our Netfinity 7000 machines has the hot-swap SCSI arrays wired backwards compared to all the others. This, in turn, enabled me to learn all about recovery procedures for disk arrays. I am very happy that I had chosen to set up the machine in question as a RAID-5 array.

Someone on a mailing list was whining about how his totally legit copy of Windows XP had taken offense at some changes he made to his system, and he’d had to phone up Microsoft and grovel for a new activation code to enter before he could boot his PC and get his data back.

Well, I’m kinda unsympathetic towards people who choose to support Microsoft by buying their software but then whine about how awful it is. I replied with a flame, pointing out that it’s perfectly possible to live a rewarding life without ever buying anything from Microsoft. I said that since he’d chosen to be Bill’s bitch in spite of abundant magazine articles about the evils of XP, he should just shut up, bend over, and enjoy what he’d paid for.

This turned into a surprisingly insightful discussion about operating system choice, during which he admitted that the main reason he was still using Windows was that he was familiar with it after twenty years, and didn’t want to have to spend money and learn something new.

At this point, I realized something: I am unlike a lot of other people, in that I mostly embrace change rather than fearing it. When CD came out, I took a trip up to London to the one store that had CD players, and listened to one. When I got home, I started saving my money. I didn’t buy a single vinyl LP after 1985.

Similarly, I embraced MiniDisc, in spite of my investment in cassette tapes. I switched to DVD as soon as all the movie studios were releasing discs, even though I had a VCR. I’m still working on switching from serial ports and SCSI to USB and Firewire. And I just switched to OS X, even though it has meant spending about $150 on software upgrades and crossgrades, learning a new UI, and spending a couple of days rearranging my hard drive and getting comfortable with how everything is best stored under the new OS. (And yes, I use the new Finder, and the dock.)

Change is good. Change is life. The opposites of change are stasis, death, conservatism, nostalgia, Republicans, and various other evils.

Sure, you have to be wary of change-for-the-sake-of-change. But when you can see that there’s a better way of doing things, surely it’s foolish to give in to inertia?

Some people may wonder why my web site was left unchanged for over a year. Well, I’m engaged in a lengthy project to digitize my entire photo collection, using a Nikon film scanner to produce 3000×2000 scans direct from the negatives.

Some of the images are decades old, and often the film has deteriorated and needs careful restoration. Color film in the 70s really wasn’t very stable, and these negatives haven’t been particularly well cared for either. My plan is to scan them, fix them, and archive them onto digital media.

Of course, this requires some care—five years ago, it might have seemed like a sensible idea to archive onto Syquest cartridges, after all. Who’s to say what will be around in another decade? Will we even be able to read most of today’s file formats? (How many art programs read NeoPaint files?)

A lot of people use TIFF. Few of them realize it, but TIFF is a really ugly file format originated by Microsoft. I say it’s ugly because I’ve read the specification. It has a zillion variations, including different byte ordering on different platforms. I’ve seen graphics packages which both claim to read and write TIFF, but won’t read each other’s files. So for archiving, TIFF is a definite no-no.

PNG is an open standard, it’s lossless, and it gets better compression than practically every comparable format, including TIFF. Because it uses no patented algorithms, it’s likely that every graphics program will at least have code to read it. Because Open Source implementations of the algorithm are available, I know that if the worst comes to the worst I can always write my own program to read PNG and write it into whatever’s the appropriate format in ten years’ time. So it’s PNG for me.

Anyway, after months of work my hard drive was getting dangerously full, so this weekend I bought a CD burner. Of all the data storage media out there, I think CD is the one most likely to still be readable in a couple of decades. I’m planning on using the Kodak pro-grade gold CD-Rs, which have a rated life of 100 years.

CD is a bit of a bitch to use, however. You have to burn the discs, verify them afterwards, and so on. On PCs, this generally involves a lot of dicking around with flaky driver software; the ThinkPad at work refuses to boot if the CD burner is plugged in, so you have to boot first, then plug and pray, and about half the time it’ll then recognize the drive. Assuming that worked, you can then try and burn a disc, which works about 80% of the time. The rest of the time the CD burning software hangs while updating the catalog at the end of the burn, and you have another coaster.

I was determined not to have similar experiences at home. Of course, I have a Mac at home, so that was a good start. Then I picked out a CD burner which was Firewire, so (a) I wouldn’t have any buffer underrun problems, and (b) I wouldn’t have to dick around with SCSI or USB drivers and termination problems.

Next, I narrowed my selection down to CD burners which were approved by Adaptec (who now want to be called Roxio), who make the Toast software used by practically everyone who burns CDs for a living.

Finally, I picked a drive which had the latest BurnProof technology. This is a hardware feature where if the drive stops receiving data fast enough—say, because Internet Explorer chokes while you’re browsing the web—the laser stops in a controlled fashion, marking how far it had got so it can continue when the data flow resumes. Which means fewer coasters, and the option of burning CDs while doing something else.

That’s the theory. Of course, no matter how careful and prepared you are, the universe has a way of screwing you over. In this case, I managed to get a faulty CD burner, and wasted most of yesterday trying to coax it into working properly. It would act just like it was working, but the CD would never verify and would be full of random (but sonically interesting) flipped bits.

Fortunately, I foresaw even this eventuality. Rather than trying to save $50 by buying online, I had decided to slum it and buy from CompUSA. So instead of paying two sets of shipping charges and waiting several days for a replacement, I picked up another burner this morning. The new one works fine. Rips at 40x, burns at 14x. Sweet!

I’ll carry on using DVD-RAM for day-to-day stuff, as it’s just vastly more convenient than CD. But now when everything’s finalized and annotated and cataloged, I can burn it on gold for keeps.

The CDRW drive I picked was a QPS Que! and in spite of the initial problems, I’m happy with it on balance.