WebOS “Post Mortem”?

WebOS Nation unsurprisingly doesn’t think so. Still, I have a sad feeling that this is one of those cases where bad PR trumps all. A LOT of people linked to that NYT article, so now whether they realize it or not, future reviewers will be forced to grade WebOS as “alive or dead” instead of baselessly and objectively.

Android benefits a little from this because its tablets are graded on the Android scale (whose 1-10 is 1-4 on the iPad scale). Sadly, since it won’t properly be weighted, the best grade a hypothetical new TouchPad with WebOS can hope to get is “alive”, maybe even “alive and well”. But that won’t let customers know that on a scale where the iPad 2 is a 9, the iPad 1 a 7, the Fire a 3 and the Xoom a 2, a new TouchPad that’s a 6 will just be labeled “not dead”.

This nonsense is a self-fulfilling prophecy. I hope it doesn’t actually play out like this.


OK, now that that’s out of the way, I have to disagree that hardware, marketing, and Sprint were at fault while WebOS itself was “perfect”.

All iPhones have underclocked processors and even sometimes bench slower than other phones with similar clock speeds. This is not something you can ding the pre for because EVERYONE does it. If every processor ran at the speed it’s engineered to, you’d have two problems. First, yield would be too low. Remember when intel had 3.0, 2.66, and 2.0 GHz dual core xeons for Mac Pros? They only fabricated 3GHz chips, which came at a certain yield. Those that failed testing at 3.0 GHz were again tested at 2.66 GHz and if passed were sold as such, and those failures were tested at 2.0 GHz and sold as such and the rest were destroyed or sold at even low speeds. That means if you bought a 2.0 GHz xeon and thought you were so clever for clocking it up to 3.0 GHz you’d probably find out why it failed. The same is true for the processor in the Pre. Yes you could set the speed back to what it was “supposed” to be but that often made the phone quite hot and battery quite dead (the second problem). I’m not sure if today’s ARM chips do, but the (mobile) G4 could run at half clock and automatically spin up to full clock while you were rendering or something.

Even though webOS devices had a GPU and could run OpenGL content on it, the UI was not run on the GPU. This made scrolling choppy, things not anti-aliased enough, and contributed to the aforementioned heat problem. This is a really hard ding because Mac OS X has been doing this since 10.2 (mid 2002). iOS always rendered everything on the GPU. While it’s true that Android didn’t either (and also scrolled like shit) the Android UI didn’t have as many animations to notice were choppy. Since webkit was GPU accelerated (CSS3 transitions) it’s ignorant of the NYT article to say webkit wasn’t ready.

Despite Palm and Sprint’s terrible marketing, the thing was hyped as an iPhone killer regardless. To say this set the bar impossibly high is just to outright admit iOS can’t be toppled. When Apple entered the smartphone business, they knew they would be measured against the best: at the time that was Blackberry “email phones”. They took a key feature, browsing, which was awful on every mobile device, and made it 10,000% better. “It’s not the mobile web, or the kinda-sorta-looks-like-the web, it’s just The Web”. Palm then took Apple’s webkit, didn’t implement GPU acceleration, lost buttery smooth scrolling and zooming, and focused on multitasking, which everyone was certain the iPhone couldn’t do.

To this day I have to keep explaining to people that iOS always multitasked (how else would an OS work) and today it still doesn’t like they thing it does. Unless granted special privileges, background apps freeze, sleep, and thaw when resumed. Perhaps because they didn’t have this method to steal yet, Palm made literal multitasking. So if you had a browser tab with a heavy AJAX and JSON loop or something, it would just keep looping while you checked your email, tried to play Angry Birds, whatever. Again, killing battery life and making the thing hot.

Saying the hardware was inadequate to do any of that basically declares that webOS 1.0 wasn’t a mobile OS and Palm should’ve shipped it on netbooks or even desktops – which is something I’d like to see today though. I maintain my belief that webOS would be best on full sized touch surfaces for buying train tickets / viewing timetables or interactive mall directories.