On 24/01/2008, Phil Wilson <[EMAIL PROTECTED]> wrote: > As far as I know, Apache cannot stream files.
The client does the streaming. Apache does the serving. Technically possible with any HTTP server. Adam Wrote: > Apache is okay, but no security and it can only do http, It has very few bugs and vulnerabilities for it's size. It can also do HTTPS. It also drops privileges once it has bound to the listening port (I think). I would have thought that would be considered some security. Phil wrote: > >> Red5 .. > >> VLC > > Even if these were OK, do they work on the massive scale required by the BBC? I haven't done stress tests on them. But if they are different streams then it's more a case of duplication and whether the OS can handle requests fast enough. > I'm going to stick my neck out and suggest that neither of these > servers are capable of handling this load as-is. Guess work or actual fact? > An RTMP client needs to have an execution environment. What precisely needs to be in this execution environment? Do I actually need to write a full Flash environment, that seems a bit excessive for just streaming a file wouldn't you agree? I think what I was planning is a little out of my league if it involves writing a Flash implementation in Java. Please reconsider using sensible formats. > <questions about dates> > See the JavaScript Date documentation for your favourite implementation, such > as > http://developer.mozilla.org/en/docs/Core_JavaScript_1.5_Reference:Global_Objects:Date Still doesn't explain how midnight is handled or what the timezone is! Upon closer inspection I notice the Beeb is carefully setting it's times one minute before the hour and one minute after. Thus I doubt the Midnight condition is ever going to happen. Well done! Timezone wise it seems to be irrelevant for the Flash player as the Server sets the time it generated the page. Appears to be in GMT. Do I have to wait till summer to find out if it switches to BST or whether it stays on GMT? Of course I could parse the dateNow field as well but my intention would be to cache the page to avoid repeated page loads. > <acceptable characters in PID and Token> > These are generated by the BBC, so you probably don't need to know, It would help to be able to sanity check what I parse. If I "mis parse" something I wouldn't want to request stupid pages would I? I guess I could just guess where it is sensible and issue a warning to the user. Would be nice to have a definitive answer, these IDs are being generated somewhere so someone must know what set they are generated from. > I would hope that most of your other questions become redundant if an API > appears, > as has been suggested. As would I, but no API yet so the answers would be good. I Wrote: > I therefore assumed that RTMP could still be used but wasn't > the recommended approach. I may have been wrong though. Typographical Error. I thought HTTP but somehow typed RTMP. It should of course read: "I therefore assumed that HTTP could still be used" Andy -- Computers are like air conditioners. Both stop working, if you open windows. -- Adam Heath - Sent via the backstage.bbc.co.uk discussion group. To unsubscribe, please visit http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html. Unofficial list archive: http://www.mail-archive.com/[email protected]/
