Third Thing

3. I remember! The third thing is that along with how the blog is now separated into two departments instead of five, I changed the method that the blog entries are reposted onto the compendium. Instead of having an email client always running that re-sends the posts to Blogger (as described in Me? Blogging?), I’ve had to change things around a bit. With the transition to New Blogger that allowed me to consolidate the blogs, something broke with how Blogger parses special characters. My department blogs have an em dash in their names—that thing. The problem is, you can’t find an em dash on a keyboard; it’s a special, high-ASCII character like • or æ or . Unlike the lower-128 characters (everything you actually see on a keyboard: letters, numbers, normal special characters like $%&@), there’s no standard ASCII code for the upper-128 characters, which, on the Mac, includes the em dash (option-shift-hyphen), but on other systems may not. So HTML threw its hands up and instituted what’s called HTML Entities instead, where special characters are represented by codes like — and • (which result in — and •, respectively). The problem in this case was that Blogger started escaping these codes, so the department links at the bottom of the posts on the main page started looking funky. Instead of sending "Genius/Idiot—Current Thoughts" (embedded in a link, of course), it sent "Genius/Idiot&—Current Thoughts" which resulted, instead of Genius/Idiot—Current Thoughts, in Genius/Idiot—Current Thoughts. Not what I wanted.

Of course, the easy thing to do here would just be to give up and change the department blog names, to something like Current Thoughts or Genius/Idiot--Current Thoughts. But I’m not well known for doing the easy thing; I wanted it pretty. Instead, I spent many long days over a period of several months—most of last year, really—learning enough UNIX to do this by hand on my own computer (the Mac is now based on a UNIX operating system, in case you didn’t know). In the end, I had to go to an incredible rigamarole in order to save my silly em dash. Here’s the new setup (this replaces step 6 in Me? Blogging?):

6. Blogger emails the post to an email account I have set up on my own computer (I have a static domain name provided by staticcling.org). It comes in, is handed off to Procmail for processing, which hands it off to formail to modify the header so that Blogger would take it back (took forever to figure that bit out), then hands it off again to sed for lots of reasonably complicated text processing, to make it look like it used to look before Blogger broke it (Blogger had changed some style stuff as well). Procmail then sends it to Postfix, a UNIX mail server, which shoots it back off to Blogger.

This was at least as hard as it sounds to figure out how to do. I could never have done it without the Procmail Quick Start and a nifty program called Postfix Enabler. I’m cheap; I fought with .conf files and a bunch of stuff I don’t even remember anymore (God help me when I need to change computers; I have no idea how I did all this) for days before I finally gave up and paid the $10 for Postfix Enabler (a Mac front-end for Postfix). I was a fool. What I failed to do in two or three days, Postfix Enabler did in two or three clicks. I love the Mac ☺.

Leopard=Windows?

Screenshots

There are some very interesting “screenshots” that have recently been released purporting to be of the next version of Mac OS X, 10.5 “Leopard.” The most tantalizing thing about these supposed screenshots is that apparently, Windows applications (in this case, Internet Explorer 7) can run natively on a Mac under OS X in Leopard.

If true, this is revolutionary. Windows and OS X applications running concurrently on a Mac? This is the Holy Grail of computing. Now, I don’t know if these screenshots are real or not. If they’re fake, they’re superbly done. But here, it doesn’t matter; I just want to talk about the idea, not whether or not Apple is actually doing it.

Explanation

Now, there are positives and negatives to this idea. Before we go into them, let’s examine exactly what we’re talking about here. The new Macintoshes (as of 2006) are now based on Intel processors instead of the old IBM/Motorola/Freescale PowerPC processors. Since Intel (or Intel-compatible) processors power all Windows (and Linux, for that matter) PC’s, that introduces a possible level of compatibility between Macs and PCs impossible previously. Already, Apple has released software called Boot Camp that allows the new Intel Macs to boot into Windows XP. Now, this is a separate boot situation: You can turn on your computer and have it be a Windows PC, or turn it on and have it be a Mac. While this is useful (for more details see my previous post on the subject) for occasionally running Windows-only software like games, it’s anything but seamless, and there’s almost no real benefit besides saving desk space over just buying an actual PC. The recently released Parallels software is another option for running Windows on your Mac: It provides an environment similar to the old Virtual PC, where Windows, and Windows applications, run in a window on your Mac. This is better than a dual-boot situation; you may lose a tiny bit of speed, but not much, because Parallels on an Intel Mac is not an emulator like Virtual PC on a PowerPC Mac; it’s a “virtualization machine” and therefore runs at near-native speed. The problem with it is that it’s still not seamless. Parallels is one application on your Mac; all your Windows applications run within that application, in a window with the Windows desktop in it. Functional, but ugly, and a bit of a pain to work with.

The ideal solution is something called a “compatibility layer.” This will allow Windows applications to exist side-by-side with Mac applications—completely seamlessly. Done right, the only way you’ll know which kind of application you’re running is by how it (the application itself) looks and behaves. Instead of being like having a Windows machine on your Mac, it would be like simply running Windows applications in the same way you run Mac applications. In a perfect world, Windows apps would exist on your hard drive right next to your Mac apps and documents and files, with the only distinguishable difference being in the icon. Mac OS 9 (Classic) applications work exactly like this on PowerPC-based OS X machines now. There is currently no way to do this, but the Darwine project is working on it, and this is what is promised by the Leopard screenshots mentioned above.

Consequences

What are the ups and downs of this last method? Well, the ups are obvious. Being able to run any Windows application natively on my Mac without having to deal with the horrid Windows operating system is, as mentioned above, the Holy Grail of computing. There have been many times where some service or game or function that I wanted to access or use was only available for Windows, and I didn’t have a Windows machine or emulator, so I and my beloved Mac were left out in the cold.

The downs are a little more interesting. Viruses are obviously the biggest threat. I don’t need to describe here how horrible the virus situation is in the Windows world. Running Windows on your Mac obviously exposes you to virus risks that are currently nonexistent for OS X. Dual booting is no more or less risky than simply using a Windows box. Your Mac is a Windows box then. The situation is similar running virtualization software; whatever partition of your hard drive is dedicated to Windows is vulnerable to Windows viruses. The virus risk for compatibility layers is an unknown; we’ve never seen one in the wild, so it’s hard to tell. There’s reason to hope, for solutions like Darwine, that the virus risk would be somewhat lessened, as you’re running Windows applications, but not Windows itself. With the hypothetical Leopard version, however, it doesn’t look like that would apply, as the screenshots imply that Windows is running in the background (just like Mac OS 9 does for Classic now). It could even increase your Mac’s exposure to viruses if, as I suggest above, Windows applications reside on the same logical drive that your Mac applications do…which is why it won’t be done that way.

But there’s a much more important potential “down,” that I mentioned in detail in my previous post on the subject: That the ability to run Windows software on your Mac will serve as a serious disincentive for developers to write new software on the Mac. This was my biggest fear before, and is echoed by others, for instance this comment on MacRumors: “[Running Windows apps natively]= the end of native Mac development as we know it”

I certainly understand why people might think so, but I no longer do. See, my Economics classes have finally started to have some effect in my brain, and I think the process will work itself out quite differently from the "Those Macies can just fire up Windows if they need to use our software. Ha ha ha (evil laugh).” scenario. In fact, given the insights from my Economics classes, I suspect it might be just the opposite: The ability to seamlessly run Windows apps on the Mac will attract millions (yes, millions) of new Mac users. This will increase the Mac’s market-, user-, and mind-share dramatically. These new converts from Windows will run their old Windows software, sure, but as time goes on, they will gradually migrate to Mac OS X applications (exactly as happened during the transition from OS 9 to OS X via Classic), because of the greater esthetic value, interoperability, compatibility and functionality of Mac software on the Mac platform vs. Windows software on the Mac platform. Besides (and this is really the killer point), it doesn’t matter if they migrate or not. Maybe they will all keep using the old software they’ve got until it’s so old that it’s useless. Still, when they go to buy new software, they will look for Mac software first. If they can’t find any at wherever they’re looking, sure, they’ll buy Windows software and use that. No big loss. The point is, though, that a developer that offers a Mac version of their software has an opportunity to make a sale that the developer of Windows-only software will miss out on. This will provide a powerful incentive for software developers to program for the Mac. No, this won’t cause every single Windows publisher to put out a Mac version. Not by a long shot. But, if Leopard does include native Windows support, and if that in fact causes a boom of Mac switcher sales, expect the amount of Mac software (and, possibly, even Mac-only software) to increase, not decrease.

Gavin Shearer of Microsoft has an interesting article with a similar perspective on this issue.

Macintel????

So Apple has decided to ditch IBM and ally with Intel. What's the world coming to? Has Hell frozen over? It's going to be a tough transition, regardless.

I'm not sure I have any new wisdom to add to what is said at the pages linked above, except to make a couple of minor points. (A lot of this is laid out very nicely in an (as usual) excellent Ars Technica article and accompanying discussion forum.)

• Forget running OS X on Non-Apple PCs. It's just not going to happen, folks. At least not in a sanctioned way. Despite what some people have said in the past, Apple is not an OS developer that happens to sell computers. Apple's bottom line is and always has been mostly CPU sales.

• The idea of being able to run Windows natively on a Mac is a neat idea, but ultimately mostly useless. Presuming the technical difficulties (motherboard differences, hard drive format, etc.) are overcome, all you've gained is a little cash and a little desk space (and, admittedly, a prettier office) over just buying a PC. And you lose the advantage of being able to use both at once.

• What excites me more is the prospect of a really good Windows emulator on my Mac. Since no processor emulation is necessary, Windows apps should positively hum, making buying a separate PC pretty much unnecessary, even, hopefully, for games. And if you just have to have the latest and greatest game and it won't run fast enough under emulation, well, that's when you fall back on the dual boot model mentioned above. This might just work out really nicely. Especially for PowerBooks. Imagine being able to haul around one machine and run any program, play any game...this might be fun.

• Neat as all this is, this is still not the CHRP platform we were promised 'lo these many years ago. As I recall, CHRP was supposed to allow the running of multiple OS's simultaneously. But perhaps I'm misremembering. I do recall having great hopes for the CHRP platform, and am still sorry it died. Consistency has not exactly been Apple's strong suit. I could probably write a whole website devoted to promising technologies Apple has abandoned (oh, OpenDoc, I miss you so!).

• I do have a couple of fears, however. As well as the Mac has been doing lately, I worry that Apple has been slowly watering down the distinctiveness of the Mac platform. For years now, Apple has been doing little things here and there to make the Mac more PC-like. You can trace it as far back as Apple changing floppy drive vendors so that the drive no longer sucked your disk out of your hand when you inserted it. OS X has several PC-like features in the interface, not least the requirement that all filenames have a damned TLE on the end. And now Macs will actually have Intel Inside. At what point will consumers decide that it's not worth paying a couple of hundred extra for a PC with a prettier case? It's been really nice (especially since the G5 was released) gloating to all my PC friends about how much faster, particularly megahertz for megahertz, Macs are than PCs. I don't like having my gloating turned back on me. Could this trend lead to the homogenization—nay, the commoditization—of the PC industry? The prospect doesn't frighten me as much as it once did. Admittedly almost entirely thanks to Apple, Microsoft seems to have released its first decent GUI OS ever—Windows XP Pro. I've used it (though not extensively), and it's not half bad. Not good enough, no, but it seems consumers are finally starting to demand something resembling elegance in their mainstream OS's. Add that to the downright cool-looking boxes companies like Alienware are putting out, and though I would still be crushed if Apple vanished or sold out, I would no longer see it as the end of the (computing) world.

• My biggest fear, though, was summed up by phjones on MacFixIt:

"Whilst I think Steve Jobs and the Apple crew would only act in the best
interests of Mac users, I still have a knot in the pit of my stomach. From my
point of view, the big question is whether Macintel machines will be able to
run Windows at full speed. If they do, it's the beginning of the end for MacOS.
At the moment, software producers have an incentive to produce MacOS-
compatible software because it gives them access to a market that would be
otherwise unavailable - admittedly some companies feel the market is too
small but that's their decision. If Macintel machines are capable of running
Windows, there will be absolutely no incentive for new companies to produce
MacOS versions of their software: "Those Macies can just fire up Windows if
they need to use our software. Ha ha ha (evil laugh)." Inevitably, less new
software will be written for MacOS and existing software will slowly drift
away.

I hope this doesn't happen."

So do I.

• Lastly—there's something that has been bugging me for years. When the PowerPC processor first came out, basically the entire computer industry was saying that CISC technology was dead. Intel was going to be able to crank out maybe a generation or two more by cramming circuits a little tighter and running a little hotter, but eventually was going to be forced to switch to RISC like the PowerPC or die. This apparently didn't happen. Does anyone know what actually did happen? Let us know.