Isn't it what these guys have done ?
But I think that what most people (I can only guess) here are looking for is a cheap, commercial, rendez-vous only service for P2P(with no fallback) without having to manage servers (like LCCS was). But for some reason this thing is not going commercial and probably never will.
sure, you have a couple of different open source implementations
but most people see that as "not stable", "complicated', etc.
what I'm asking/proposing here is a bit different
- do you want something easy to install server/client side ?
- do you want something cross platform windows/linux/osx ?
- do you want it cheap ? or at least cheaper that what you can actually find ?
- do you want it hosted for you ?
- do you want it open source ?
- do you want commercial support ?
if it interest nobody I don;t really care
I will just implement for my own needs
and not planing at all for others to use it
developing such server just for me
or for an amount of people with a different amount of experience/knowledge
is not the same thing
I ask because the thing is compatible in term of development with another project I'm already working on
but "just for the fun of it" I'm not gonna build a full product if nobody is interested
At the time I was looking for exactly what cirrus proposed, but for commercial use. So Hosted (and obviously commercial/payed for).
The only other option was LCCS. But a constant connection had to be maintained to the server and that had a cost per hour.
I never really understood why a connection had to be maintained to the server after the P2P link had been established. Maybe it is unavoidable.
I didn't and still don't have the skills to setup and maintain a linux server and I believe it to be the case for most people. The only options available for doing P2P only with no fallback are hosted AMS services which are overkill just to do peer introduction.
I have a feeling a commercial version of cirrus would be popular, but then again not many people seem to be raising their hands here.
I certainly would be interested in such a service if affordable, but I've given up hope for that some time ago.
yep no problem
it is as simple as that
I'm working on redtamarin
redtamarin - AS3 running on the command-line - Google Project Hosting
a highly customised version of the open source AVM2 (also known as Tamarin, avmplus)
so while I'm working on that, and now that the RTMFP is documented to the W3C
I can basically build a command line RTMFP server (eg. I got the runtime I just need to write the software)
and so I'm testing the water to see how many ppl would be interested by that
that is all
only difference it would make is get that kind of RTFMP server sooner than later
as personally don't have a huge need for it for now
> so while I'm working on that, and now that the RTMFP is documented to the W3C
note that RTMFP is documented in the IETF, not the W3C.
yes and thank you for the astonishing work
without such spec I could not even dream to start implementing a RTMFP server
I know I would.
Something that troubles me these past few days, is that I don't know when/if Adobe will close the curtains on Cirrus at some point.. rendering my whole Cirrus based games disabled.
So, having control over the server that is running Cirrus (or something like Cirrus) would be a big+ without any doubt.
Let me know if I misunderstood something.
you got the right idea
controlling your own server is always best
but it has also other advantages
- it can work on "closed" networks, think RTMFP inside the enterprise
- you can distribute customisable RTMFP servers to users, think minecraft server
- you can customise the server itself, think "MyServer extends RTMFPServer"
for use case where you want to log, send emails, etc. based on RTMFP "actions"