A scenario started playing through my head the other day. In the late 1990s, Apple looked dead. Then they released OS X, plus very cool shiny hardware. That put Apple back in the game and gave them the life they needed to bring about the iPod, etc. Apple’s revival didn’t come from iPods and iPhones; it came because they made a deep connection to the software developers. In 2000, if you went to a developer conference, everyone was carrying some kind of PC laptop, probably running some version of Windows, possibly Linux. But almost overnight that changed, and it changed completely. By 2003, any self-respecting developer was carrying a MacBook, preferably the one the size of a small aircraft carrier. Apple did an undeniably brilliant job of growing this beachhead among the developer community into a dominant brand. Everyone wanted what the cool kids had. Apple had a winning product: they had the most beautiful version of Unix ever, with a user interface that beat anything that had ever appeared on Linux or Windows.
But now, Apple is looking more and more hostile to the developer community that enabled their revival. OS X is evolving into a slightly more capable version of iOS, and we’re all dreading the day when the only way we can compile and install our own software is by using Apple’s proprietary tools and going through the App Store. If you look closely at Apple’s work, it’s clear where they’re putting the effort: there’s a race condition in basic text editing (TextEdit, Mail, etc.) that’s been around since at least OS X 10.6, and I suspect goes all the way back to 10.1. It’s not something arcane that only crops up in strange circumstances; I run into it every day (and on several different machines). And that’s only the start.
I know loads of developers who are saying, “yeah, I’m assuming I’ll be off Apple in a few years.” But that’s a problem: it’s one thing to talk about leaving Apple; it’s something completely different to know where you’re going. Developers aren’t going to run back to Microsoft; they might get somewhere with Google’s ChromeOS cloud vision (it’s a powerful vision, but I’m not yet ready to sign up for it), though that begs the question of how to do bread-and-butter enterprise backend work. The Linux community seems to have settled into unproductive infighting between Canonical and the rest of the world. Dell and HP got nothin’. But looking back at what Apple did over a decade ago: is there anyone who can put together both very cool hardware and a cool operating system for software developers?
Just before Christmas, an answer occurred to me — and it came completely out of left field. A few months ago, Valve started looking for beta testers for their new SteamBox gaming computer. (I don’t have one.) According to several teardowns, the SteamBox is flexible and expandable, a quality that’s been notably absent from recent Apple hardware. Valve has also been making noise about SteamOS, a Linux distribution tailored to running games. And then it hit me: Valve has both the vision and the technical chops needed to build really cool, high-performance hardware and package it with a winning operating system.
Could Valve do to Macintosh hardware what the Mac did to the PC market over a decade ago? I think they could. Do they want to? I have no idea. The SteamBox is currently a desktop machine, and a developer-oriented product line must include a laptop. There’s no sign that Valve is thinking about laptops, nor do they make sense in the high-performance gaming world. For that matter, a hot desktop box stuffed full of GPUs may not make much sense either, unless you’re doing a lot of data analysis. SteamOS has a cool user interface, but it’s tuned to gaming, not to software development. If you want a developer desktop, SteamOS ships with Gnome 3 Shell and Gnome Classic, neither of which represents a great leap forward in user experience. And I don’t see any sign that Valve is interested in rethinking the developer desktop.
So, yes, if you want to call this fantasy, pie-in-the-sky thinking, I’m guilty as charged. If you think the near-term future of software development will be based on some platform other than the laptop/desktop, I wish I could agree. All these “ifs” aside: if I were Apple, I would be looking over my shoulder. At Valve.