nu's not unix


How the Mac got its mojo back

posted on Mar 5, 2013

Further to my last post, Mike Arrington has some similar thoughts about the success of the Mac being directly related to the Internet.

Arrington seems to think consumers went to the Mac simply because the playing field was levelled by the Internet. But there's more to it. Regular people don't just jump platforms because the playing field is "level". It took them a long time to learn how to use their computer. They won't change without a good recommendation. And who do you ask for a recommendation? The "computer guy" you know, who is probably a programmer or IT person for something or another.

Those "computer guys" (and gals, of course) were among the early adopters of Mac OS X because they are the types who are willing to jump platforms on their own. They are the ones who aggressively kept Windows XP up to date, upgraded OS X as soon as upgrades became available, and tried out new Linux distros for fun--everything that the casual computer user was not willing to do.

These folks were the first switchers, and found much to like. A superior environment for doing your work in, good compatibility with Microsoft file formats, and great compatibility with the unix boxes in the data centre, with all the killer commercial apps available. It was, in fact, the ultimate compatibility machine--the complete opposite of what a "closed" architecture should be like. Windows soon got relegated to the home desktop rig, running games, Linux got relegated to the server, and OS X found itself doing the day-to-day work because it was the versatile, plays-nice-with-everybody notebook that got lugged everywhere. But because of that portability, it was the Mac that people saw being used, and that put it into people's heads that it was an option for their next upgrade. If the computer guy uses it, then… And so they started asking about it.

In those early years, I had this conversation with Windows users, many times:

Them: I didn't know you were a Mac guy.
Me: I'm not, really, I only starting using one recently.
Them: How do you like it?
Me: It's great, no problems at all.
Them: I heard you don't have to worry about viruses.
Me: That's true, pretty much.
Them: I need a new computer, mine sucks. Do you think I should I get a Mac?
Me: Depends. Do you use anything that is Windows-only?
Them: Oh yeah, I use Office a lot. :-(
Me: Office is available for the Mac.
Them: Really?

Or, if I was talking to someone in the Linux world:

Them: I didn't know you were a Mac guy.
Me: I'm not, really. The new Macs are unix boxes.
Them: Wut?
Me: Yeah, full BSD, like a NeXT. See, here's my terminal.
Them: What!?
Me: Yeah, they come preloaded with emacs, gcc, perl, apache, X11, all kinds of stuff.
Them: They can run Linux software?!
Me: Most open source stuff will compile. See, here's Gimp. Here's LaTeX.
Them: Whoa...
Me: Plus, it runs Office and PhotoShop.
Them: Fuck.

Done, sold. The Mac faithful never really had these same conversations, because they had always used Macs, had always put up with the differences, and people saw them as iconoclasts with somewhat suspicious opinions. But when the computer guy's ThinkPad disappeared one day to be replaced with a PowerBook, that made you think and ask questions.

That's how Apple beat both Linux and Microsoft at "open computing" -- by being open enough to run everybody's software on a single platform, and do it well. Nobody else did that. They still don't, over 10 years later.

Open does beat closed

posted on Mar 1, 2013

John Gruber missed the simplest and most obvious rebuttal to Tim Wu's "open beats closed" thesis, and that is:

Apple is the largest vendor of open source operating systems in the world.

Like many people, possibly including Gruber, Wu is under the mistaken impression that Apple is down at the "closed" end of the scale. He fundamentally misunderstands how openness really works in the world of computing. Apple does understand however, because they are secretly a rebranded Unix shop.

It may not seem like a big deal to people who were among the Mac faithful the whole time, but the early "switchers" in the OS X 10.0 and 10.1 days were hugely important to driving early adoption and development on OS X. They were mostly unix and internet developers, for whom openness is vitally important.

The Mac was a superior platform for these developers because of its openness in the respects that mattered to them--compiler toolchains, libraries, POSIX, source compatibility with other Unixes, command line environments, internet protocols, etc. The Internet and web have always been dominated by "open" unix-based systems, and the Mac almost instantly became the Web and Internet development workstation of choice because of how nicely it fit into this world, right at the exact time when the Internet was exploding into everyday use.

The Mac was actually a spectacular success in Internet workstations (especially laptops) for the exact same reasons Linux was a success in the data centre. And when I say, "exact same", I mean exact: the same command line tools, the same programming languages, the same shell scripts, the same database environments, the same web server. It was like your Linux production environment was running right there on your laptop. Just rsync and your changes were live on the real server. Yeah, you could do that from a Linux workstation, too, but setting up a decent Linux laptop in 2003 was an exercise in pain and frustration, and even if you got it working, you still didn't have access to the world of commercial apps for Mac (and, after the Intel switch, Windows). It was the PowerBook that freed the unix power developer to go wild on the Internet.

Gruber is right, the consumer doesn't really care about openness. But developers do, and Internet developers especially. And where the developers are, the consumers are bound to follow, because that's where the cool stuff happens first. Microsoft proved that in the 1990s, and it was openness that drew developers to the PC, too.

Some might argue that the transition from web to mobile means that Apple's closed attitudes with iOS will become a problem. Maybe, but I don't think so. Mobile development is still Internet development by and large, just with a different front-end. The fundamental concerns of the developers are with communications, protocols, service architectures, and "the cloud", and iOS still has those unix roots that keeps it open in the ways that matter to developers. Android has unix underpinnings, too, of course, but Apple captured the cream of the Internet developer market 10 years ago, and Android is a latecomer. They either have to woo away Mac developers who have already made their choice, or they have to woo the ones who stayed behind with Windows or Linux on their workstations, and who enjoy programming in Java to boot. I suspect this factor explains everything you need to know about why Android has the "feel" that it does.

iOS maps SUCK!!!1! No, wait, that other thing...

posted on Sep 21, 2012

I upgraded to iOS6 just to get the new maps app so I could see what all the bitching was about.

BTW, I'm a map geek. And I love Google Maps. So I was prepared for disappointment. But frankly the new iOS 6 maps are the bomb. I get it, their map data is young and immature, and there will be many, many bugs of the kind that can only really be shaken out with real world use. But I didn't actually see any of those, perhaps because I'm lucky enough to be downtown in a major city that has lots of up-to-date info available on it.

What I did see was an awesome 3-D satellite view (shown above) that could not only be panned, zoomed, and spun, but could do all of that automagically by enabling the compass to set the perspective on the map. It's basically bird's-eye augmented reality. That's hot.

I also tried the turn-by-turn directions, which is something I almost never use (having been blessed with a pretty reliable internal compass) but which I had heard bad things about. So I asked Siri for directions to the Cambie pub, to see how rough around the edges it really was. Siri couldn't find the "Candy" pub, but after I over-enunciated it, she figured it out, and gave me directions. It was a pretty simple task, since it was only 2 blocks away, but everything was pretty darned precise, down to the exact metre that I needed to turn left, and the ETA, which was spot-on. I thought Siri actually misjudged the location of the pub door, because I had to walk an extra 10 steps past it before she told me I had arrived. But then I noticed that I was right in front of the Hotel main entrance, which also serves as an entrance to the pub, so good enough, and arguably more correct.

As for the accuracy of the map details, I learned that the alley closest to my office has a name: Trounce Alley. This surprised me, because (A) I didn't know it had a name, and (B) the map showed that Trounce Alley actually extends back two blocks to a section known as Blood Alley. On Google Maps, the same 2-block stretch of alley is called Blood Alley Square. So who is right? Turns out the alley is officially named Trounce Alley, and Blood Alley Square is just a colloquial name for the far eastern end of the alley where it joins Carrall Street. So Apple (top) got it right, and Google Maps (below) gets it wrong:

Speaking of quality of data, the Google Maps above shows the Woodward building as a construction zone, which it hasn't been since 2009. The Apple Map shows a completed building, so it's data is not only more accurate, but more more current in this case.

My only complaint is that the compass is a bit finicky, complaining regularly of compass interference as if there were big magnets nearby. I regularly had to reset it by waving the phone in a figure-8, and when I set the map to auto-rotate as I turned about, it would swing from side to side a little erratically, sort of like being on a rocking boat. If boats rocked in a spinny kind of way.

I'm sure there's lots of junk data somewhere in the gargantuan data sets that go into mapping apps like this, but for the first week of an app this huge, I'd say they nailed it.