I’ve been watching HP’s latest midlife crisis implosion, wherein they are effectively killing the WebOS platform that they bought for $1.2Bn about a year ago, and trying to sell their PC business which is currently the largest in the world, to become an enterprise company, where they are merely one of several big fish. The latter madness is very much akin to the Carly-era “let’s become a Whitebox PC Vendor and sell off our differentiating technology” fail, but the former is something I’m currently interested in. I’ve been chatting about it in news discussions, some of which are publically visible, but wanted to collect thoughts here.
1. HP has done this shit over and over. Apollo in 1989. Convex in 1995. DEC via Compaq in 2002. This may be the most egregious “Buy a distressed asset with good technology, then abandon it” of all time, but the precedent was certainly there.
2. It took Apple (roughly, based on patents and employees) 22 years and two failures (the Newton and one killed internally around ’04) to build a mobile platform that didn’t suck too much for even the reality distortion field to market, and they made a good try at fumbling again at release (remember “No one needs native applications, webapps only“?). Palm was the only other contender with that kind of background, and they designed around Apple’s mistakes (Hello competent multitasking, clipboard, and notifications system!). From a design standpoint, WebOS was the obvious winner.
3. Android, WebOS and Meego are all fundamentally ARM EABI Linux, with a shiny UI coating. iOS’s XNU for ARM and POSIX-y underlayer isn’t that different. WebOS managed the balance perfectly, with official support for optware for the “This is a Linux box” software, and a commercial store for the UI-focused software. Apple kicks and screams and tries to subvert users who want to do anything other than in-band UI-wrappered software through their official store, but there is still a substantial jailbreak/cydia/etc. aftermarket. MeeGo didn’t really have a usable mobile UI layer, restarted too late, and isn’t really worth discussing. Android has technological impediments built in to it to make it as unpleasant as possible to run native binaries, even though the capability is there, which has resulted in a store full of $1.50 pieces of cross-compiled FOSS software with clunky dalvik wrappers over them, and my contempt.
5. I’d like to reiterate that the WebOS development environment was brilliant. The tool-chain was built completely out of open, standard components and languages. No custom JVM substitute to provide lock-in, patent wars, and wasted cycles. No effectively domain specific language (ObjectiveC with weird libraries? Really?). Just a clean interaction model built on top of HTML, CSS, and Javascript, native binaries in C/C++ using STL and OpenGL, and an interface to connect the two.
4. I pretty much agree with this Ars article on how it should have gone down, with an additional “they should have hurried the fuck up” to cap it off.
What I really want in a device is a pocket POSIX workstation with a clean Mobile UI. I’ve ranted about this before. A WebOS device with decent hardware (High-resolution screen, usable keyboard, removable storage) would have been nearly perfect, and the Pre3 was almost it. But it was six months late, lacked removable storage, and has now been cancelled between its European and US release. I’ll probably end up with a HTC Doubleshot (aka. MyTouch 4G Slide) shortly here, as Android is the last platform standing that meets my most basic requirements, and T-mobile and HTC (now with Unlocked Bootloaders!) are the least evil respective sources of service and devices. At least there is a Linux system buried under there somewhere, and now that Matias Duarte has moved to google the Android UI might improve. And hell, maybe I’ll bargin-bin a WebOS device in a couple months as they’re being dumped, just for sport.
