Archive for the ‘Computer Industry’ Category

How does interoperable video recover from Chrome dropping H.264?

I don’t know that it does.

It’s essentially inconceivable that the rest of the world — particularly once you move beyond the browser — is going to ever adopt WebM, so any dream of a universally interoperable video format is right out the window. The best possible outcome we can hope for now is that the major browser developers will all support WebM, and we’ll have one video standard for the web, and one for everything else. And that’s not a very good outcome at all, really, compared with the kind of ubiquitous video compatibility we could have achieved.

Even that limited victory seems pretty remote. It would require Apple to support support WebM in Safari and on its mobile devices, despite the fact that Google dropped H.264 in large part just to piss Apple off. And it would require Microsoft to support WebM, despite Google being an arch-rival. It was hard enough to get Microsoft to support H.264, a neutral open standard. What are the odds they’ll ever support WebM?

Google has just thrown away the last, best hope for truly interoperable video, and ruined web video for the foreseeable future. That might sound apocalyptic, but it’s hard to imagine how things could develop otherwise.

Why is Google dropping H.264 support?

Ditching H.264 serves two strategic purposes for Google. First, it advances a codec that’s de facto controlled by Google at the expense of a codec that is a legitimate open standard controlled by a multi-vendor governance process managed by reputable international standards bodies. (“Open source” isn’t the same thing as an “open standard”.) And second, it will slow the transition to HTML5 and away from Flash by creating more confusion about which codec to use for HTML5 video, which benefits Google by hurting Apple, since Apple doesn’t want to support Flash.

It also, of course, really sucks for consumers.

This move is, in other words, a thoroughly nasty bit of work. It’s not quite as bad as selling consumers down the river to Verizon on ‘net neutrality, but it’s close. And if Google is actually successful in making WebM, not H.264, the standard codec for web video, they’re literally going to render hundreds of billions of dollars worth of tablets, smartphones, set-top boxes, etc. with H.264 hardware support obsolete.

“But wait!”, the OSS fans are saying. “Isn’t Google really standing up for freedom and justice, because H.264 requires evil patent licensing?”

No. Expert opinion is that WebM infringes on numerous patents in the H.264 pool, and will need a licensing pool of its own to be set up, just like Microsoft’s VC-1 did. So the patents are a wash. This is Google manipulating the market entirely for selfish advantage here, and it’s all the worse because they’re pretending otherwise.

A recap of Google duplicity on H.264/WebM

Google has announced they’re dropping support for H.264 in Chrome because their own WebM is “more open”. This is, essentially, a lie. Let’s recap.

H.264 was developed by the Moving Picture Experts Group (MPEG) and the Video Coding Experts Group (VCEG), which are standards committees that draw members from industry and academia under the umbrella of the International Organization for Standardization (ISO) and International Telecommunication Union (ITU), which are (in practice) intergovernmental public/private partnerships. It is governed by a public, multi-party governance process representing many different stakeholders.

This is how a real open standard works.

Does H.264 require patent licensing? Yes. But that’s not because (as some people seem to believe) it was developed by some company that licenses it out to make money. Rather, it requires patent licensing because it turns out that a lot of the techniques you’d want to use in a modern video codec are patented, so the standard that MPEG/VCEG developed (basing their decisions largely on technical considerations) ended up infringing on a bunch of them — about 1000 patents in all, in fact. As is common with such things, a patent licensing pool was set up to make licensing all of those patents cheap and easy for implementors.

It is extremely unlikely that WebM does not infringe on some of those very same patents. It’s a very similar codec to H.264. Moreover, this has happened before. Microsoft’s VC-1 codec was supposed to be a patent-free alternative to H.264, and guess what? It ended up requiring a patent licensing pool as well.

So, with H.264 you get a codec that’s an actual open standard, with a formal multi-party governance process and with easy patent licensing. With WebM, you get a codec that’s not formally standardized, has no formal governance process (and is de facto controlled by Google, because they employ most of the developers), and that has huge ‘submarine’ patent risk.

Google, somehow, has the gall to claim that the latter is “more open”. And a lot of people seem to believe them.

Why Microsoft won’t do to tablets what they did to netbooks

A lot of early netbook models ran Linux, but over the last couple of years, Windows has emerged as dominant in the netbook market as it has traditionally been in the desktop/laptop markets. Could the same thing happen in the tablet market, once the ARM version of Windows starts shipping?

No.

It’s fairly clear that touch-based computing just doesn’t work very well with desktop user interfaces. Microsoft, however, still seems to be pretty much out to lunch about the idea that they need to provide a more appropriate UI. They’ve been selling tablets for 10 years, and they’ve made no serious attempt at this.

Microsoft’s primary strength in the market, with Windows, is application compatibility. Anything that significantly disrupts that, like requiring developers to design new UI, or even possibly just requiring them to compile for and test on a different architecture (like ARM) severely undermines the one big thing going for Windows.

Particularly in this instance, where Windows/ARM tablets, by the time they actually make it to market, will be competing against established competitors. There are already 60,000 iPad apps. How many apps were there that were optimized specifically for Linux netbooks? Essentially none. They ran the same desktop apps as other Linux systems, which end users just don’t seem to find especially compelling relative to Windows-based offerings.

Additionally, it’s very unlikely Microsoft has the guts to do anything like requiring developers to create new user interfaces if they want their apps to run on a hypothetical Windows/ARM tablet. Now, the geek brigade might not see a problem with giving users a ‘choice’ to run apps with non-optimized interfaces, but the real-world consequence of such permissiveness will be lots of half-assed ports that destroy the platform’s user experience.

Even worse, Microsoft probably lacks sufficient internal coordination for a full-court tablet push, if such a push requires the redesign of major Microsoft products across multiple divisions. See here, for instance.

When we were building the tablet PC in 2001, the vice president in charge of Office at the time decided he didn’t like the concept. The tablet required a stylus, and he much preferred keyboards to pens and thought our efforts doomed. To guarantee they were, he refused to modify the popular Office applications to work properly with the tablet.

OK, I assume Microsoft has gotten rid of that guy by now, but such problems appear endemic to Microsoft’s organizational culture, which is why you see Microsoft still chasing off in several different directions in the mobile space.

A Microsoft that was likely to be a serious competitor in the tablet space would have already announced a tablet operating system based on Windows Phone 7, and they’d have the confidence to back that system as their sole tablet platform. The Microsoft that exists in our world is still clinging to the notion of running desktop versions of Windows on tablets (despite a decade of market rejection — with this ARM announcement, it seems they’ve decided that was all Intel’s fault), and when it eventually does occur to them to do a tablet version of Windows Phone 7, they’re unlikely to have the level of commitment necessary to deliver the product they’d need to deliver to be relevant.

Of course if Ballmer goes, this whole assessment is right out the window. There are clearly a lot of talented people working for Microsoft. Microsoft’s vast resources, under the right leadership, could put Microsoft back in the tablet/smartphone game pretty fast.

Balmer Talks Tablets

See here.

Microsoft’s behavior here is entirely bizarre. They’ve been pushing desktop versions of Windows on tablets with absolutely no mass-market success for over a decade. Apple releases a tablet with a much more lightweight OS, with UI redesigned from the ground up for a touch device, and has instant mass market success.

And Microsoft’s announces that they’re taking this very seriously… and their response will be to keep pushing desktop versions of Windows on tablets, but… what? Hope that people notice this time because Apple got people paying attention to the tablet market?

Right, good luck with that. The reason people didn’t pay attention to those Windows tablets is because they sucked, and they’ll be perceived even more negatively now that there are better products on the market to compare them with.

Here’s how I see this going: some time in the next 18 months, Microsoft will finally come to terms with the fact that they’re barking up the wrong tree trying to stick a desktop version of Windows onto a tablet. They won’t actually abandon that approach, because internal corporate politics won’t let them and they lack the discipline to actually focus on things that matter. But they will announce an additional project to create a tablet versions of Windows Phone 7, which might even turn out to be a credible technical effort.

Only problem is, by the time that gets market, it’s going to be going up against the iPad, Android tablets, WebOS tablets, and the newly announced Blackberry tablet, some or all of which will be quite well established by then and will be on their second or third generation. And few or none of the value network effects Microsoft benefits from in the Windows market will do anything for them in the tablet market. So, basically, they’re screwed.

What Google is up to with WebM/VP8

There are three reasons to introduce an alternative to H.264:

  • It’s technically better.
  • It doesn’t require patent licensing.
  • You want more control over web video than the multivendor H.264 standardization process provides.

VP8 isn’t technically better than H.264, and it will almost certainly have the sort of patent problems that will require a licensing pool to be set up, just as with Microsoft’s VC-1, another attempt at a patent-free H.264 alternative.

That means what Google is after is control.

Google has played the PR game masterfully here. The release of VP8 is being hailed as a victory for openness. Meanwhile, if VP8 were to actually displace H.264, then instead of a codec controlled through an ISO standardization process involving dozens of vendors, you would have a codec controlled almost entirely by Google.

Advocates of open computing need to think really, really hard about what’s going on here. A lot of them seem to have declared for Google (and against Apple) because of a couple of fairly trivial technical points with no long-term strategic relevance (Android runs apps that aren’t from its app store, etc.), and some clever rhetoric on Google’s part. But the paradigm that Google wants for the future of computing — cloud-based apps running on Google’s servers — is no more compatible with openness than Apple’s vision of appliance-like computing devices.

In fact, it’s less compatible. Apple’s devices can at least be jailbroken; the cloud can’t be. And Apple doesn’t seem interested in having access to all of your data, whereas Google’s cloud-based vision for computing would involve them having access to all of your data, and Google’s business model, based around targeted advertising, creates an incentive for them to know as much about you as possible.

Google’s release of VP8 codec changes little

I’ve previously discussed the dustup over Ogg Theora vs. H.264. Today Google announced it is releasing the VP8 codec as open source. VP8 is an ostensibly patent-free codec which, unlike Theora, is said to be as good, technically, as H.264. But it won’t change much.

Most of the case against Theora didn’t come down to its technical quality, but the fact that H.264 is a standard that extends far beyond just web video on desktop platforms, and there was no way Theora plausibly had a chance in that larger world. The same logic applies to VP8.

The Mozilla Foundation and Google can implement VP8 in Firefox or Chrome or whatever. But let’s look at, say, Netflix. Right now, they can serve the same H.264 streams up to desktop platforms via Silverlight, to iPads (soon iPhones and iPod Touches), to the PS3, to the Nintendo Wii (as of this month), to Roku boxes, to the Xbox, to various Blu-ray players….

Most of those devices implement H.264 decoding in hardware, and their vendors couldn’t add VP8 decoding to them if they wanted to. And they mostly don’t want to, because H.264 licensing is pretty cheap, and they don’t care about the ideological arguments against it. It uses the same kind of licensing regime used by MPEG-2 (DVD), Audio CD, ATSC digital broadcast… the industry is very comfortable with this model.

So, what is a company like Netflix supposed to do? Keep two copies of its entire library, one in VP8, for Firefox, and one in H.264, for every other client platform in the world? Because the Mozilla Foundation has some ideological objections to software patents, and so refuses to implement H.264?

Right. Good luck with all that.

Had VP8 been around (and open source) five or six years ago, things might have been very different. But H.264 has already achieved critical mass. It simply won’t be displaced by another current-generation codec. If people want a patent-free codec to eventually emerge on top, they should start thinking about how to build a patent-free next generation codec, and get it ready for prime time before H.264′s inevitable MPEG-LA-licensed successor is ready.

Adobe’s JooJoo reaction validates Apple’s flash stance

Engadget’s review of the JooJoo Linux tablet:

But what about Flash? This is supposed to be the big differentiator, right? The iPad killer! In an interesting move, Fusion Garage coupled the Atom processor with NVIDIA’s Ion graphics to aid in playing full screen Flash video (or for doing… something). Unfortunately, the software just isn’t there yet. Currently the device is running Flash 10.1 beta 1, and won’t have hardware-accelerated Flash video for a good while now (the timing is partly reliant on Adobe support, and is labelled as a “work in progress” by JooJoo). That means some regular-sized YouTube and Hulu works, as decoded by the CPU, but full screen Hulu is jittery, and a 720p YouTube clip is like watching a slideshow. In one of the biggest moves of irony, JooJoo has actually implemented a hack for YouTube where you can view a video in Flash or in “JooJoo” mode which is a straight playback of the MPEG video file every YouTube video harbors. What does this remind us of? HTML 5, albeit with a less elegant implementation. This of course only works on YouTube right now, though JooJoo says it plans on supporting other sites in the future.

They’ll probably eventually get hardware acceleration working and get this straightened out, but there’s a larger point here. Platform vendors shouldn’t have to rely on the good graces of an outside company to make key elements of their user experience not suck. Even if this gets fixed on Linux/Atom tablets, it will only crop up again with some other new platform in the future.

The notion of a key piece of web infrastructure being proprietary software controlled by a single vendor was only sustainable within the homogeneity of the traditional desktop world. Now that computing is seriously moving beyond the desktop — now that the industry is actually interesting again — it’s all falling apart for Flash. Apple is just hurrying things along a bit.

A conversation that will occur this summer

Normal Person: [switching instantly back and forth between Evernote and Safari while listening to a Pandora stream, with Loopt updating their location continuously in the background]

Geek: “Hey, you know your iPhone doesn’t support ‘real’ multitasking.”

Normal Person: “Huh?”

Why the iPad is a big deal

In terms of long-term large-scale impact on consumer computing, the iPad is the most significant new computing platform to launch in over 20 years, with the possible exception of the Web, depending on your definitions. Really. Even if the iPad itself ultimately ends up as a minor player in the tablet market, it will likely be the device that takes tablets mainstream, like as the Mac did for the GUI. Why is this a big deal? It’s a big deal because this form factor has been anticipated for over 40 years both in fiction (see, for instance, 2001) and by human/computer interaction designers (see, for instance, the Dynabook), and will likely have a major impact on the shape of computing over the next couple of decades.

In terms of how users interact with computers, the iPad is the most significant thing since the mouse and the modern GUI. You could credit the iPhone for the multitouch revolution, of course, but I think the the phone form factor is just too limited to develop the full vocabulary required for touch-based UI to really come into its own. The iPad is the first platform that will enable that.

And there’s even a bit more to the iPad than that. The iPad isn’t just a new device built along the same model as traditional computers but with a new form factor and primary interaction mechanism. Put together the simple model lineup, the lack of focus on geeky tech specs, the elimination of the file system as a user-level concern, the system-level automatic application installation, removal, and updating… and the iPad is also clearly a major push toward the sort of appliance-like personal computing that many have tried for over the years, but none have previously achieved.

Now, I understand perfectly well that these kinds of predictions are easy to dismiss. That’s sort of the nature of the thing; if this were all obvious, there wouldn’t be much point in writing about it. But large chunks of the web get archived, and we’ll see in five or ten years just who saw what was coming, and who didn’t.