Getting started with TS3 SDK (ts3_sdk and ts3client_pluginsdk)

[poktomi]

Hi guys,

I'm a newbie to development on TS. My background is in "traditional" SIP-based VoIP development.
I would like to integrate Teamspeak with an existing SIP platform and have started by downloading Teamspeak 3 SDK an client plugin SDK.
I have to admit however that I do not quite know where to start.

I am using Visual Studio 2005 on Windows Vista and
- running the example/client in the SDK results in an exit code -1 in SLEEP(300) of all things!
- cannot even compile the plugin SDK plugin.c as it chokes on:
1>c:...\ts3client_pluginsdk-3.0.0-beta11\ts3client_pluginsdk\src\plugin.h(21) : error C2143: Syntaxfehler: Es fehlt ')' vor '('

In general: where do I find meaningful information on how to code specifically:

- how to create a TS "session" (i.e. logon to a server with a given identity)
- send audio
- receive audio
- connect to a counterparty
- receive a handle to audio stream and remote identity on incoming "call"

Thanks a lot in advance,

poktomi
etc.

[PeterS]

Compiling the plugin SDK: Our project files are for Visual Studio 2008. I added another one for Visual Studio 2005 for the next plugin SDK release, perhaps that solves your issue. Also I tweaked the paths in the project file to use relative paths, so users don't need to adjust them anymore to their machine.

exit -1 in the SDK client sample: I have no idea. The sample runs fine on all our test machines.

> how to create a TS "session" (i.e. logon to a server with a given identity)
See chapter "Connecting to a server" in the included documentation

> send audio
This is done automatically when voice input was detected. You don't have to call a specific function to send voice data. But vice versa, you can prevent sending voice data by muting the input device.

> receive audio
Same as above. If playback devices were initialized, any incoming voice data is automatically played, except the output devices were muted.

See the chapter "Using playback and capture modes and devices" in the documentation about how to initialize input and output devices.

> connect to a counterparty
> receive a handle to audio stream and remote identity on incoming "call"

The TeamSpeak 3 architecture works differently than classical SIP. You don't connect to a counterparty, you connect to a server and the server handles all the transmission, user identities etc. The SDK samples should give you the basic information for starting, anything further is in the documentation.

[poktomi]

Compiling the plugin SDK: Our project files are for Visual Studio 2008. I added another one for Visual Studio 2005 for the next plugin SDK release, perhaps that solves your issue. Also I tweaked the paths in the project file to use relative paths, so users don't need to adjust them anymore to their machine.


exit -1 in the SDK client sample: I have no idea. The sample runs fine on all our test machines.

Hmmm, ok. should I try on Linux and see if the same happens?

> how to create a TS "session" (i.e. logon to a server with a given identity)
See chapter "Connecting to a server" in the included documentation

> send audio
This is done automatically when voice input was detected. You don't have to call a specific function to send voice data. But vice versa, you can prevent sending voice data by muting the input device.

> receive audio
Same as above. If playback devices were initialized, any incoming voice data is automatically played, except the output devices were muted.

See the chapter "Using playback and capture modes and devices" in the documentation about how to initialize input and output devices.

> connect to a counterparty
> receive a handle to audio stream and remote identity on incoming "call"

The TeamSpeak 3 architecture works differently than classical SIP. You don't connect to a counterparty, you connect to a server and the server handles all the transmission, user identities etc. The SDK samples should give you the basic information for starting, anything further is in the documentation.

First of all thanks for the detailed reply.
Regarding the IDE: When will that release (VisStd05) be out and where will it be announced?
In your experience: what's better to develop on Linux/debian or Windows.
I can use both and if Linux is easier I can just use that.


Unfortunately the docu assumes you know the insides of TS. Once I fully understand the interaction model the exact methods, classes etc. will be of use. Until then what I'm looking for is a sample which I can use as a skeleton to expand on. Basically a commented "Hello World". Some context..

Thanks again,

-p

[poktomi]

Hi Peter,

is there any way we can get in direct contact (e.g. can I send you a private message?) That would probably get me started quite fast.
You will find that only my first questions seem stupid, once I get a basic overview we will soon start to talk seriously...

thanks,

(ich wohne in Berlin und bin auch gelegentlich in München)

-p

Compiling the plugin SDK: Our project files are for Visual Studio 2008. I added another one for Visual Studio 2005 for the next plugin SDK release, perhaps that solves your issue. Also I tweaked the paths in the project file to use relative paths, so users don't need to adjust them anymore to their machine.

exit -1 in the SDK client sample: I have no idea. The sample runs fine on all our test machines.

> how to create a TS "session" (i.e. logon to a server with a given identity)
See chapter "Connecting to a server" in the included documentation

> send audio
This is done automatically when voice input was detected. You don't have to call a specific function to send voice data. But vice versa, you can prevent sending voice data by muting the input device.

> receive audio
Same as above. If playback devices were initialized, any incoming voice data is automatically played, except the output devices were muted.

See the chapter "Using playback and capture modes and devices" in the documentation about how to initialize input and output devices.

> connect to a counterparty
> receive a handle to audio stream and remote identity on incoming "call"

The TeamSpeak 3 architecture works differently than classical SIP. You don't connect to a counterparty, you connect to a server and the server handles all the transmission, user identities etc. The SDK samples should give you the basic information for starting, anything further is in the documentation.