X
Tech

Is openness an 'inherent advantage'?

The discussion about Google dropping support for H.264 produced some comments that complaints about WebM ignored the inherent advantage and goodness of openness - as well as suggestions that given that WebM was originally developed as the proprietary, commercial VP8 codec and hasn't been submitted for formal standardisation, calling it more open than an open but commercial standard really depends on your definition of open.
Written by Simon Bisson, Contributor and  Mary Branscombe, Contributor

The discussion about Google dropping support for H.264 produced some comments that complaints about WebM ignored the inherent advantage and goodness of openness - as well as suggestions that given that WebM was originally developed as the proprietary, commercial VP8 codec and hasn't been submitted for formal standardisation, calling it more open than an open but commercial standard really depends on your definition of open. (Leaving aside all the important questions of whether WebM breaks any patents, something that Google decided remarkably quickly for so fraught a legal area.)

Increasingly, open is used as a synonym for free - which has its own overloaded range of meanings. Free software always used to be, as we said 'free as in speech, not free as in beer' - and free speech is philosophically virtuous but not always advantageous. Free speech allows people of a political opinion different to you to speak their opinions however much you dislike them; without drawing invidious comparisons, free speech can allow people to say things most people disapprove of. For businesses, there's also the problem of 'free as in puppy'; a free puppy brings high costs in puppy food, dog walking, floor cleaning and dog sitting when you take a holiday, and free software comes with support costs (pay Red Hat or pay the cost of growing internal experts). Open used to mean that you could find out about the standard - not that the standard was free of patents and royalty charges; conflating open standards with open source software is the Humpty Dumpty approach of making the words mean what you want them to mean.

Is the fact that the software is 'free' or 'open' enough to outweigh any potential disadvantages? Like disliking royalties and patents, that isn't a technology question; it's a philosophical question. And the value of openness isn't a technology issue; it's a philosophical one. Technology can be open and good or open and bad; it can be good because it's open or in spite of it (and it can be good or bad, because of or in spite of being proprietary, commercial, patent-encumbered technology). It's kind of like mandating organic, hormone-free, locally grown food; not everyone wants that. Freedom of choice also means freedom to choose things not everyone approves of... (And if you're planning to tell me that open software X is better than closed software Y, please include specific features, specifications and measurements rather than generic criticisms of closed software Y to keep the discussion qualitative.)

As to the technology question: is WebM better than H.264? Although Google promised at Google IO last year that there would be hardware acceleration for WebM, it's not there yet. It doesn't do lossless compression (which is an option in H.264). When Google bought it, On2's video test set was only 70 videos , which engineering manager Jim Bankoski called a wide test set (I've been told the H.264 test set is more like thousands). It didn’t use the SSE2 or SSE3 multimedia acceleration instructions in x86 chips or even multiple cores - so it was only faster than non-hardware accelerated H.264 on older machines and on Snapdragon ARM processors 640x40 video at 30fps was using 100% of the CPU. Bankoski called the best quality mode "pretty slow" and mentioned several improvements in video quality that were only in internal builds without giving a date for when they'd be in public versions. I haven't yet seen a side-by-side comparison that convinces me of the WebM advantage - though I'd be delighted if I do. It will give IE 9 users more choice than other browsers that have picked one or the other; back in May the IE 9 team said "IE9 will support playback of H.264 video as well as VP8 video when the user has installed a VP8 codec on Windows". That's an odd comparison to be making with Google.

Mary Branscombe

P.S

There is a precedent for all of this, I think. Android's media player doesn't support WAV, either of reasons of openness or because Google didn't want to pay the royalties. That means if your voicemail arrives as WAV files of a particular flavour (which mine does), you have to redirect it through a cloud service or find a WAV player in Android Marketplace to listen to it on your phone. If I've just paid $BIGNUM for a shiny new phone, I'm not always happy to shell out even another dollar for a feature I've had on every other smartphone for years for free.

Editorial standards