OSNews: http://www.osnews.com/story/21810/Apple_Proposes_HTTP_Streaming_Feature_As_Protocol_Standard Exploring the Future of Computing en-us Copyright 2001-2009, David Adams adam+nospam@osnews.com Tue, 08 Dec 2009 14:25:16 GMT http://www.osnews.com/images/osnews.gif OSNews.com http://www.osnews.com Doesn't it already work? http://osnews.com/thread?372863 http://osnews.com/thread?372863 Can't we already stream a video using HTTP? I mean, you can embed a video, use a source URL like http://www.example.com/movie.wmv or http://www.example.com/movie.flv and the file plays as soon as it begins loading... Could someone clarify, please? Fri, 10 Jul 2009 17:54:00 GMT donotreply@osnews.com (Liquidator) Comments RE: Doesn't it already work? http://osnews.com/thread?372866 http://osnews.com/thread?372866 Apple's scheme is to chop a video file into a series of MPEG2 transport files, perhaps at every 10 seconds. The main video file is a M3U file that points at all of the segments, and a streaming client will use HTTP to request a series of files. This lets the client use a fairly primitive method to start a stream at a given time offset, avoiding some of the complex issues with HTTP byte-range streaming.

It's not really that complicated, but it does offer some capabilities beyond just playing a stream from a single HTTP connection. Fri, 10 Jul 2009 18:17:00 GMT donotreply@osnews.com (unwiredben) Comments Comments from Roy Fielding (HTTP spec author) http://osnews.com/thread?372867 http://osnews.com/thread?372867 Interesting comments from Roy Fielding Fri, 10 Jul 2009 18:34:00 GMT donotreply@osnews.com (twanj) Comments RE: Doesn't it already work? http://osnews.com/thread?372915 http://osnews.com/thread?372915
Can't we already stream a video using HTTP? I mean, you can embed a video, use a source URL like http://www.example.com/movie.wmv or http://www.example.com/movie.flv and the file plays as soon as it begins loading... Could someone clarify, please?


that's not live streaming but video on demand. live streaming has been implemented by real, quicktime, wmp & flash with the real time streaming protocol which need special streaming servers. http streaming can use normal http-servers and needs less processing power on the server.

ms' smooth streaming for silverlight does this and this company

http://www.internetlivetv.eu/english/info.html

implements the idea with flash. apples implementation seems to be the only one with an open protocol which doesn't need one specific plugin for the webbrowser. looks like the perfect streaming solution for html5 video.Edited 2009-07-11 01:55 UTC Sat, 11 Jul 2009 01:54:00 GMT donotreply@osnews.com (puenktchen) Comments RTSP issue? http://osnews.com/thread?372970 http://osnews.com/thread?372970
The biggest issue with RTSP is that the protocol or its necessary ports may be blocked by routers or firewall settings, preventing a device from accessing the stream.


I am not sure what the hell Apple was thinking when they stated that this is somehow an issue with RTSP. OF COURSE RTSP can be blocked on firewalls and/or routers: it is kind of the point of it! If bandwidth is not a concern, I am pretty sure that most network administrators would be glad to let people use it freely on their networks (assuming that it will not raise other concerns, such as lost productivity on the work place among others).

What Apple is proposing will lead to additional costs for most networks as packet filters with rules based on ports/protocols will no longer be effective against excessive bandwidth usage and people will have to deploy firewalls/trafic shapers/whatever that check every incoming packet on port 80 to find out if it is regular HTTP packet or if it is this new thing that they came up with and deal with it acoording to their policies.

Having said that, anything that does not require any proprietary streaming solution and the likes is always a plus in my book so more power to them, I guess! Sat, 11 Jul 2009 16:28:00 GMT donotreply@osnews.com (DeadFishMan) Comments RE: Comments from Roy Fielding (HTTP spec author) http://osnews.com/thread?372989 http://osnews.com/thread?372989
Interesting comments from Roy Fielding



Jealously doesn't serve Roy well.

http://www.ietf.org/mail-archive/web/ietf/current/msg57133.html Edited 2009-07-12 01:34 UTC Sun, 12 Jul 2009 01:34:00 GMT donotreply@osnews.com (tyrione) Comments RE: RTSP issue? http://osnews.com/thread?373001 http://osnews.com/thread?373001
I am not sure what the hell Apple was thinking when they stated that this is somehow an issue with RTSP. OF COURSE RTSP can be blocked on firewalls and/or routers: it is kind of the point of it!


but apple doesn't provide the band-width, they sell the iphone. from their perspective, the blocking of rtsp sucks. Sun, 12 Jul 2009 11:31:00 GMT donotreply@osnews.com (puenktchen) Comments RE[2]: Comments from Roy Fielding (HTTP spec author) http://osnews.com/thread?373086 http://osnews.com/thread?373086 I hardly think he's going to be jealous about what's described in that 'standard'. Hell, the mime type they're using hasn't even been registered. Mon, 13 Jul 2009 10:06:00 GMT donotreply@osnews.com (segedunum) Comments Comment by TBPrince http://osnews.com/thread?373088 http://osnews.com/thread?373088 Well, everything which will become a standard extension is welcome. Bandwidth concerns are, IMO, quite weak because everybody is "streaming" via HTTP nowadays. Microsoft also provides extensions to IIS7 to ease streaming contents via HTTP.

However, as someone else noted, this is not true "streaming" but rather video-on-demand because actual streaming go far beyond that.

Anyway, I'm mixed about this approach because that would require splitting your videos and mantaining them. That would mean extra work for webmasters and potentially a mess when a lot of videos get uploaded on a server. Plus, everybody is able to record a video by their phone and publish it while not everybody could be able to split files, build playlists and so on.

On the other hand, this approach would allow browsers to work better with audio/video. Hmmm.... Mon, 13 Jul 2009 10:15:00 GMT donotreply@osnews.com (TBPrince) Comments RE[2]: Comments from Roy Fielding (HTTP spec author) http://osnews.com/thread?373095 http://osnews.com/thread?373095
"Interesting comments from Roy Fielding



Jealously doesn't serve Roy well.

http://www.ietf.org/mail-archive/web/ietf/current/msg57133.html "
tyrione, I can't see what a comment on a comment on a comment has to say about the original comment author - especially when it doesn't reference the original comment at all. Moreover I don't see anything in the any of the comments which suggests that the original author is jealous.

Am I missing something? Did you post the wrong URL? Mon, 13 Jul 2009 11:40:00 GMT donotreply@osnews.com (steogede2) Comments