So Microsoft technology will support an open video format, while Google's Chrome only supports proprietary ones? Clearly, this is a confusion/FUD move by Microsoft.
Erm, Chrome supports Theora in HTML5 video. I just played a few of the examples from http://double.co.nz/video_test/ as well as the Opera page linked from there to test, and they worked fine.
This is, technically, an effort by people to ensure Theora is usable in most web browsers. In other words, they're bringing a decoder to IE and any that don't support Theora ... if the user has Silverlight/Mono.
Not to burst any bubbles, but Silverlight isn't a plugin that everyone has in the way that Flash is. Silverlight seems to be at slightly less than half of all users (http://www.riastats.com/). So, that would mean anyone with Firefox, Chrome, or Silverlight could watch it and that would be pretty awesome since I think Microsoft will push for greater adoption. But right now, Silverlight isn't so well adopted. Better than the situation was before the announcement, but still not up to Flash's penetration (and well under 94% of traffic).
The grandparent appears to be assuming that all IE users have silverlight, when only 40% of them (about a quarter of all users) have it.
But you in turn forget that there is a similar Java plugin that does the same thing and is installed with Mac OS X by default. All combined I'd say 95% is a fair number so maybe the grandparent was aware of this option and simply didn't mention it explicitly.
Vorbis audio, which is playable in later versions of Flash, must be as close to 100% of desktops as you're ever going to get.
You're conflating HTML5 support and plug-in support. For instance: that Flash supports Vorbis does not mean that Flash-supporting clients can play <audio src="foo.oga">. It just means that the site can choose to deliver <audio> to one client and <embed> to another. The same situation happens with h.264.
Which goes to say: this is a good thing for Theora, but it is not a victory over h.264. Even in the most optimistic estimate, including this kind of plugin-as-equivalent conflation, it would be a tie, and only in the single domain of playback in desktop browsers.
Thanks for the link -- it looks like I was drastically overestimating Silverlight's market penetration.
However, the graphs on that site appear unreliable -- for example, they show Silverlight's market share dropping by 14% between January 22 and 26. Do you know if the raw data is available?
Travis keeps the sites confidential, for obvious reasons. It's plausible that there was another shift in sites sampled recently, similar to the doubling that it reported for Silverlight starting one week in November:
http://www.flickr.com/photos/jdowdell/4180222368/
I saw Alp's post last night... plausible, a good engineering challenge. Not sure how it will play out in the world though. But it sure added some spice to this morning's Theora-vs-H.264 wars.... ;-)
Except that most modern mobile devices DO have a hardware h264 decoder, but aren't fast enough to run theora as a plugin -- if they even have support for plugins.
It's a matter of battery life, not performance. Every modern mobile device has a sufficiently powerful CPU to decode Theora in real time, but decoding with dedicated hardware (regardless of codec) uses less power.
That's a (common) partial misconception. It is possible to get at least partially accelerated Theora decoding on mobile devices, now. For example http://www.schleef.org/blog/2009/11/11/theora-on-ti-c64x-dsp... has some writing on a decoder for OMAP3 chips (Moto Droid, Palm Pre).
That isn't quite the same as H264 in hardware, but should have comparable performance and more importantly should improve battery life considerably.
Your "94%" figure does not account for users on Windows XP, which does have Silverlight pre-installed. Mozilla has said that 60% of their users are running Windows XP, so one imagines the share for Internet Explorer is likewise significant.
21 comments
[ 4.1 ms ] story [ 63.5 ms ] threadI see a win/win here, if it's opensourced!
* http://en.wikipedia.org/wiki/Usage_share_of_web_browsers
But you in turn forget that there is a similar Java plugin that does the same thing and is installed with Mac OS X by default. All combined I'd say 95% is a fair number so maybe the grandparent was aware of this option and simply didn't mention it explicitly.
Vorbis audio, which is playable in later versions of Flash, must be as close to 100% of desktops as you're ever going to get.
Which goes to say: this is a good thing for Theora, but it is not a victory over h.264. Even in the most optimistic estimate, including this kind of plugin-as-equivalent conflation, it would be a tie, and only in the single domain of playback in desktop browsers.
However, the graphs on that site appear unreliable -- for example, they show Silverlight's market share dropping by 14% between January 22 and 26. Do you know if the raw data is available?
I saw Alp's post last night... plausible, a good engineering challenge. Not sure how it will play out in the world though. But it sure added some spice to this morning's Theora-vs-H.264 wars.... ;-)
That isn't quite the same as H264 in hardware, but should have comparable performance and more importantly should improve battery life considerably.