Just wanted one last visit to this topic. FLV wrapped H264 would be very nice, I can live without
the VP6, but the FLV wrapper gives me some extra options for our system. Foremost of which
is commercial insertion directly into the stream without using playlists. Playlists can be bypassed
and affect our revenue stream.
We really like the Roku box, but without this (which should be possible since it’s only a wrapper) we are
stuck with going after the boxee market which I don’t really care for that much.
I’m not positive, but it sounds like you might be talking about Flash’s real-time streaming (RTMP), which is independent of the FLV container format, and a whole different animal.
I have rolled my own server side streaming system that manipulates the video stream
on the fly. After watching the process of people writing hulu plugins that circumvent
hulu’s commercials, then hulu breaking their plugin, and the process starting all over again,
I decided to look into how I could put out streams that can’t be used without our ads in them
(or optionally without ads if a subscription is payed).
We patented the method last year and are using it for testing in our XBMC
plugin. If you would like to see it work I will send you the xbmc plugin. I have
tested this both with the VP6 and H264 codecs inside FLV wrappers and it
works flawlessly.
Our system isn’t currently built to parse out the mp4’s. That’s on our list of upgrades
for the future. But until I reach that point I don’t know for sure the same method will work.
Not trying to be a smart ass, but Roku wants to sell boxes, people want content, so if more
content can be provided by adding a feature, I would think Roku would be all for it. I don’t
know how difficult it would be for you, I just hoped that Roku would consider it.
“AVP777” wrote:
Our system isn’t currently built to parse out the mp4’s. That’s on our list of upgrades
for the future. But until I reach that point I don’t know for sure the same method will work.
Not trying to be a smart ass, but Roku wants to sell boxes, people want content, so if more
content can be provided by adding a feature, I would think Roku would be all for it. I don’t
know how difficult it would be for you, I just hoped that Roku would consider it.
We would certainly consider it if we were getting asked for it. Fact is, pretty much everyone is using h.264 either in mp4 or via HTTP Live Streaming. I’ve heard from may two partners in the last 6 months for whom not having flash was an insurmountable problem. With everyone targeting non-flash-enabled devices (iOS and others), they’re investing in migrating to other technologies.
Yes, if nobody is using FLV, perhaps it’s because it is unsupported and undocumented. So, rather than twiddling our thumbs waiting, hoping for support, we’ll just get started on converting to what is already working, h264. The poll was from last march, and no movement was taken on it at all, so we’ve given up and submit to your preferred format. with all due respect, of course. At this point it shouldn’t make much difference as long as the existing codecs continue to work.
“destruk” wrote:
Yes, if nobody is using FLV, perhaps it’s because it is unsupported and undocumented. So, rather than twiddling our thumbs waiting, hoping for support, we’ll just get started on converting to what is already working, h264. The poll was from last march, and no movement was taken on it at all, so we’ve given up and submit to your preferred format.
What a bizarre response… As you pointed out, the poll was posted last March, and after nearly a year, a total of three developers responded affirmatively (the rest of the posts were from the non-developers with generic, non-specific, pro-additional formats comments). What about that would suggest that Roku should focus any efforts on supporting the FLV container? Without RTMP streaming support, FLV container support seems pretty pointless to me.
If it was me, I’d want as many codecs and containers to be supported as possible, to make it easy for more content to be available. By requiring one of two formats to be used for any and all content (yes I know others ‘may work’ - but who has the time to test?) - perhaps that could be why it’s taken this long to get more than three developers making channels? Is that more bizarre?
“destruk” wrote:
If it was me, I’d want as many codecs and containers to be supported as possible, to make it easy for more content to be available. By requiring one of two formats to be used for any and all content (yes I know others ‘may work’ - but who has the time to test?) - perhaps that could be why it’s taken this long to get more than three developers making channels? Is that more bizarre?
With well over 100 public channels and countless private channels, I think your “three developers” math may be a little off. They put the question out to the developer community to see if there was any interest. With only three developers saying yes, it makes perfect sense to me that they would focus their efforts elsewhere. If you don’t have five minutes to test whether your FLV videos work or not, PM me the URL of one of your FLV videos and I’ll test it out for you. It will literally take me longer to type this response than it will for me to test whether your video works or not.
TheEndless, I’m just saying to post any kind of poll on the developer forum isn’t going to give an accurate sampling to make any kind of decision over. With a forum I help run, we have 30,000 registered users, and our last poll, stickied, stuck on the front page for 2 weeks straight 24/7 labelled as important, please give input, got a total of 14 responses. When you are getting 200 unique visits per day, * 14 days, that isn’t a good sampling either.
When you register, most websites get an email address - email the poll or question out and it might provide you with better information, or modify the roku box itself to run a poll once a month on a single question for all the users out there. Simply pointing and saying three people responded to a post on a forum in support of FLV isn’t anywhere near an accurate method to eliminate anything.
If the poll actually had a few options like - yes, I would use FLV if it was available, or no, I don’t care about FLV - at least then you could probably point to it and have more clout in your final decision, saying of those who responded 3 said they want the feature and 10 said they couldn’t care less. Still, it wouldn’t be as accurate as an even unbiased sample. Monetarily, perhaps those three developers who wanted FLV could have created channels to rival Netflix and Hulu, but we’ll never know now because they left to develop for other platforms.
“destruk” wrote:
TheEndless, I’m just saying to post any kind of poll on the developer forum isn’t going to give an accurate sampling to make any kind of decision over. With a forum I help run, we have 30,000 registered users, and our last poll, stickied, stuck on the front page for 2 weeks straight 24/7 labelled as important, please give input, got a total of 14 responses. When you are getting 200 unique visits per day, * 14 days, that isn’t a good sampling either.
When you register, most websites get an email address - email the poll or question out and it might provide you with better information, or modify the roku box itself to run a poll once a month on a single question for all the users out there. Simply pointing and saying three people responded to a post on a forum in support of FLV isn’t anywhere near an accurate method to eliminate anything.
If the poll actually had a few options like - yes, I would use FLV if it was available, or no, I don’t care about FLV - at least then you could probably point to it and have more clout in your final decision, saying of those who responded 3 said they want the feature and 10 said they couldn’t care less. Still, it wouldn’t be as accurate as an even unbiased sample. Monetarily, perhaps those three developers who wanted FLV could have created channels to rival Netflix and Hulu, but we’ll never know now because they left to develop for other platforms.
destruk,
While you do have a valid point, I think it would be helpful for you to remember that Roku is a small company with very finite resources. There are quite a few items that should be addressed but Roku uses it’s resources to fill the biggest needs. Squashing bugs that are making current channels unstable must be a high priority. Adding features that have been requested by many users for a long time (like official USB support) must be higher priority than adding official FLV support when there hasn’t been a demonstrated demand. Would some people use it if it was offered? Sure. Would Roku add FLV support if they had 5 times the number of full time developers? Probably.
I can understand that. Just the little nagging issues keep gnawing at me, and I can get passionate about what I think needs to be done is all. No offense is intended.
