On ter?a-feira, 29 de novembro de 2016 18:01:48 PST Max Kholmyansky wrote:
> Hi,
> 
> 2 basic questions:
> 
> 1. What's the exact difference between SECURED=1 and SECURED=0 compiled
> library versions?
> Is "1" a superset of  "0"? Implementing a different behavior? What exactly
> "0" cannot do?

SECURED=0 is a debug mode. It disables a lot of functionality, including 
encryption, just so you can develop your application.

Such application must never be deployed. It will never pass OCF certification.

What's more, a SECURED=0 application will not interoperate with SECURED=1 
applications. In order for you to test an application with real devices (when 
those exist), you'll need SECURED=1

> 2. If a server resource is created as "secure" (with OC_SECURE flag):
> I understand it means that "it can be accessed in both non-secure and
> secure (DTLS) way" - right?

The other way around: it can only be accessed in secured way. If the flag is 
absent, then IoTivity allows it to be accessed without DTLS.

Note that allowing access outside of DTLS is out of scope of the OCF 
specification. I don't remember whether that will also cause a certification 
failure. Either way, you should not have devices without the OC_SECURE flag.

> If this is the case, how does the IoTivity client determine whether or not
> to use encryption?

I think you should not ask the question.

Instead, use encryption, period.

-- 
Thiago Macieira - thiago.macieira (AT) intel.com
  Software Architect - Intel Open Source Technology Center

Reply via email to