[hpsdr] The UDP ports in the new Ethernet protocol violate the ethos of the Internet.

[Michael Cozzi]

     Simon,

     ok, so I know you are apparently one of the important folks in the 
community, so I'm pretty shy about being critical.

     But if there isn't a VPN involved, the data should never leave the 
local network. In fact, I would hope the TCP/IP stack on these radios is 
limited to communication with specific IP rangers or subnets.

     Let's not even get started on IP6.

     So I don't think the port is the issue. If port 1024 was used, it 
should never be out on the Internet. And if a VPN is involved it should 
not be hosted on the radio (what a nightmare that would be!).

     So the port is irrelevant. Data sent over the WAN, assuming proper 
VPN, would never be detectable to anyone (including the ISP).

     Unless you are suggesting that the data be sent without a VPN, in 
which case I need to find another hobby :)

     Please tell me that is not what you are doing?

     Michael KD8TUT

On 1/15/2016 2:47 AM, Simon Brown wrote:
> Port 1024 is a favourite for Trojans, virus writes and the spawn of other
> evil organisations.
>
> I do envisage issues getting out through networks not under the owner's
> control, with the new protocol supporting 48k and 96k it's realistic to
> expect the SDRs to be used from many, many kilometers away.
>
> This problem does *not* affect the ports to which the SDR sends UDP packets,
> this is fully configurable.
>
> Simon Brown GK4ELI
> http://sdr-radio.com
>
>
> -----Original Message-----
> From: Hpsdr [mailto:hpsdr-bounces at lists.openhpsdr.org] On Behalf Of Michael
> Cozzi
>
>       Well, I agree you have a point, but in practice most of the ports in
> that range are protocols which are rarely used by hams- or in some cases by
> anyone else.
>
>       Just like no one adheres to fully to 2821 anymore... many of those
> protocol assignments are dead protocols. Personally, given the application,
> those ports don't bother me. Now if it was intended to be a WAN protocol...
> then I might have issue.
>
>       But it is a good point which should be considered.
>
>       Michael- KD8TUT
>


 1452845794.0