I am frustrated with the Red5 documentation and the relevant google results. They seem to jump straight into Java programming of the server, without explaining simply how to get started with some precompiled server applications for some typical client-side programming of the Flash client, e.g. live video chat applications. I even found some blogs discussing using ActionScript2 only (assume this is only server-side, since the Flash client understands ActionScript3 although the blog authors don't make this clear either), yet I've also seen another answer on this website stating that using a scripting language on the Red5 server requires writing some Java code.
First for initial quickest start for development, I want a "wide-open" (i.e. no security against free-loaders a.k.a. "intrusion") server-side application (i.e. rtmp://host/application) that accepts all clients, live streams, and shared object connections. Shouldn't (or does) such a server-side application for Red5 exist in the official distribution? Is it as simple as copying an application file to the correct directory? Specific source and destination locations should be documented in a quick start guide, could someone tell me in an answer here?
On the flash media server, this was accomplished (my live video chat was working on both Adobe's flash media server and Wowza) with a few lines of straightforward ActionScript, given the simple and well-documented server APIs. I forgot how I accomplished it with Wowza, yet I don't remember it being very difficult. Yet with Red5, I can't seem to wrap my head about the mention of scopes (a.k.a. rooms) in the discussion of migration from Adobe's server to Red5 in the documention. It seems the API and Java class hierarchy for Red5 is either undocumented, documented in dispersed tutorials, and complex. Must I really write Java (install Eclipse, figure out how to create jars, etc) just to get this simple quick start?
At the next stage, I want security. I will need a Red5 application that exposes a server-side API to PHP (and probably Python), to allow/disallow stream and Sharedobject names.
Disallow should destroy any open so-named stream or SO.
I reckon security through obscured naming is not sufficient to implement all desired features. Will need also to include in the said API to inform the Red5 server-side application which tokens are allowed access to each stream and/or SO.
Shouldn't (or does) such a server-side security application for Red5 come standard?
