We have been talking about hUC (HTTP USRP Client) since about the time we released DVSwitch Server. So, where is it already? hUC in it's early form did not really take that long. Thought it connects to DVSwitch Server differently, It's not much different then pyUC (python USRP Client) Then we started having our usual "what if" discussions. Our wives will tell you this can go on for days. A very big issue with a web based client is security. If you expose a HTML client to the internet, people will play with it. This could be disastrous. We need security to control who can do what. I sure don't want anybody with access to a search engine to start talking on DMR TG 91 (with my call and DMR ID) So, security had to be added.
Now I have this DVSwitch Server in the cloud with a web interface and at least I can control who has access. We know DVSwitch Server is single user. It sure would be nice if I can give my friends access to my server. Let's face it, I'm not using it all the time. I can share. Some friends are more responsible then others. I would like to give Jack the ability to transmit, but not move the Server to another mode or talk group (tune). I would like to give Jill the ability to listen, but no transmit and no tune. OK, we can do that. Jack and Jill can both listen to traffic on the server. Jack can transmit. I am the only one that can tune to another TG or mode. OK, this is getting good.
We don't want to limit the fun to only web clients. What about DVSM and pyUC users? Yes, you can assign the same rights to DVSM and pyUC users. You can have users from all 3 clients with different rights. Everybody that can listen will hear not only the connected Steerable Bridge (SBridge) they will hear any client with transmit rights. Let me be CLEAR, This is one SBridge shared among multiple users. This is NOT DVAAS (Digital Voice as a service)
We did not forget AllStar. You can have one or more AllStar nodes directly connected to the Server. AllStar users (by node) can be assigned rights. RX, TX Tune. It's up to the Server owner how these connections are controlled. Some Server users use AllStar to mix audio from multiple modes. This is usually done by bridging each mode into a seperate AllStar "private node" This give the Server operator the ability to connect and disconnect modes into the main AllStar node. He can also directly connect 2 modes together without connecting those modes to anything else. AllStar is VERY versatile that way. In the case of unlike modes DMR (AMBE) to P25 (IMBE) this works quite well.If you are connecting like modes YSFn (AMBE) to NXDN (AMBE) this leave something to be desired. This is transcoding where none is required. It's up to the Server operator to decide if this is a disadvantage. No matter if it's like modes or not, using AllStar to bridge Digital Voice modes has one downside, You loose the metadata that is transmitted along with the voice. Some people don't care, the advantages of mixing in AllStar outweighs the disadvantages. I can make a case for either choice.
So, how do we give people another option. That option is called Analog_Reflector. Consider a piece of software tha t will accept multiple connections from clients (DVSM, pyUC or hUC) it will also accept connections from multiple AllStar nodes either on the same server or a remote server. (via USRP) You can have multiple bridges connected. These bridges are fixed by mode and TG (reflector). Last but not lease is a Steerable Bridge (SBridge) This is the bridge you can change modes, you can change TGs it's exactly like the bridge in DVSwitch Server now. Any client (with the proper rights) can steer the bridge. This includes The AllStar nodes. One big advantage, no more macros needed (but still supported) The command path between AllStar and the reflector is two way. We can send commands for AllStar to the reflector (and on to the SBridge) We can also receive commands from a connected client and send them to AllStar. If you want to connect your AllStar node to another AllStar node, you can do it from a connected client. AllStar becomes "just another mode" accessible from a connected client. (Think Web tranceiver)
One last point I may have glossed over. EVERYBODY get the metadata. Client, Bridges, AllStar and SBridges. Everybody gets the data. Not everybody can do anything with the metadata (AllStar) but AllStar does send what little metadata it knows about the connected and transmitting stations into the reflector.
Over the next few months we are going to build a demonstration bridge. we will show you how to connect like modes without transcoding. We will show you how to send traffic to multiple modes at the same time. An example of this is DMR to YSF AND P25. We are going to do this with one instance of MMDVM_Bridge and a transcoder. Everybody gets the audio, everybody gets the metadata. Yes, it's all done with UDP ports. It's all modular. Yes it can get complex. (we have some thoughts in that area too)
DVSwitch is about giving you control. You control the mode, you control the destination. You control the metadata. It's your server. Do what YOU want.. There is SO much more I did not cover (as Mike will tell me) There is a admin interface for Analog_Reflector (and a dashboard) Somebody "sitting on his mike"? Kick him or mute him. Somebody abusing your server, ban him. I won't promise you we have thought of everything. That will come to light as we see tAnalog_Reflector in the field.
73, Steve N4IRS