YouTube Explains Where HTML5 Video Fails 426
awjr writes "YouTube have pretty much come down on the side of Flash having major issues with the lack of features that the HTML5 <video> tag has and may never have."
This discussion has been archived.
No new comments can be posted.
YouTube Explains Where HTML5 Video Fails
Related Links Top of the: day, week, month.
Real Users never know what they want, but they always know when your program doesn't deliver it.
Re:Without content protection, we would not ... (Score:2, Interesting)
Eeew.
the worst thing is that I click it and it WITHOUT PERMISSION gets information from my PayPal account. I only use paypal with throaway debit cards, but it should not have that kind of information!
Re:Stop raining on our OSS parade with your "facts (Score:3, Interesting)
HTTP Streaming vs. RTMP. Yes, you can stream with HTTP on Flash, but for long-format video or live streaming, HTTP doesn't really have what it takes.
Re:Stop raining on our OSS parade with your "facts (Score:2, Interesting)
"Er... Are you implying that making a video available through a tag is somehow harder than through a flash app ?"
It is. It's more effort because you need to sniff for user agents and then decide either which browsers to support or not, or create different content for different clients depending on which codec they support. On the other hand we have Flash which is basically guaranteed to work as-is in over 90% of clients. I'd call that the easier choice.
Re:Stop raining on our OSS parade with your "facts (Score:3, Interesting)
I don't have any studies to cite. Safari, Chrome, and Firefox have all started running plugins in a separate process. This is primarily due to Flash taking down the browser repeatedly.
Safari even added a special error message to keep people from hating the browser:
http://fukamachi.org/wp/wp-content/photo/misc/flash_crash.png [fukamachi.org]
Before this was supported, I personally checked out the stack traces on most Safari and Firefox crashes. It was almost always executing a flash function.
When it's not crashing, it's draining a lot of battery. I suspect this is because it simply loops instead of sleeping for events.
Re:Stop raining on our OSS parade with your "facts (Score:3, Interesting)
Free browsers will never be able to support H264
That is entirely incorrect. Most OSes provide APIs for playing h.264 video, which free browsers can easily use.
Mozilla is just choosing to not take advantage of this offer.
Re:Stop raining on our OSS parade with your "facts (Score:4, Interesting)
Re:Stop raining on our OSS parade with your "facts (Score:1, Interesting)
Unfortunately flash is also guaranteed to leak memory in about the same number of cases. As well as causing a host of other problems. While HTML5 might not be the panacea some have wished for, there is a reason many are wishing for just such a thing, and we shouldn't lose sight of that fact. And a decent framework can take care of most (but not all) of the problems you mentioned.
Re:Stop raining on our OSS parade with your "facts (Score:5, Interesting)
actually, you don't need to do any user agent sniffing at all.
Here is what you do. take your video tag in a common format, use it. If it fails, by web standards, the tag is supposed to default to the contents of the tag instead.
Put a video tag inside that in a format supported on other browsers. If it fails, by web standards, the tag is supposed to default to the contents of the tag instead.
Put your flash video object inside of that.
Flash DRM (Score:4, Interesting)
With one big exception: DRM (or as the article calls it "Content Protection"). While I don't think it's impossible, I think it's a pretty big effort to produce DRM that content owners (like the MPAA or RIAA) are satisfied with as an open standard. I think they perceive open standards to be inherently insecure (despite several cases of the opposite like OpenSSL).
And in fact it's the exact opposite :
Flash's DRM is a stupid joke - in short the key to decode the encrypted RTPME streams is a a couple of filestats of the ".swf" player application, i.e.: something publicly available. No password or crypto key involved (for a longer description, look for a mirror of RTMPDump [lkcl.net]). So there's no real encryption happening and as such, Flash' DRM might even not be covered by the DMCA or local clones.
HTTP's Authentication or Session and/or HTTPS provide already enough content protection at the hosting/serving level of the video. No need to add more DRM shit on the player level.
Re:Stop raining on our OSS parade with your "facts (Score:4, Interesting)
Also worth mentioning, is that Google acquired YouTube in 2006, and Google is a supporter of Open Source [google.com] with an open source operating system [blogspot.com]. If they did look at this from an outside, objective perspective, I trust Google will do anything they can to speed up HTML5 video support.
But Google has sided with Adobe in their spat against Apple, and YouTube has a lot invested in DRM, at the behest of the media cartel. That DRM is included in Adobe products, and not in the html5 spec. That's an internal conflict for Google, and in the "principles VS revenue" conflicts, the principles rarely win.
Re:F. YEAH SEEKING: set currentTime (Score:3, Interesting)
But I turn JavaScript off. As annoying as Flash is, it's confined. JavaScript is almost always worse.
Re:Stop raining on our OSS parade with your "facts (Score:3, Interesting)
Re:Stop raining on our OSS parade with your "facts (Score:3, Interesting)