[OpenTRV-dev] Christmas secure protocol joy!

[Damon Hart-Davis]

Hi,

I’m going to get the protocol running as-is, but I am happy to take alternative suggestions (and being able to discard things not understood that are not yet defined is part of the aim) and corrections particularly over the next few days.

Even after we do an initial roll-out there is still scope to fix the protocol, though it would be nicer not to have yet another ‘legacy’ one floating around!

I’m only not working on the protocol because I’m working with our valve samples back from Shenzhen!

Rgds

Damon


> On 29 Dec 2015, at 13:24, Damon Hart-Davis <dhd at exnet.com> wrote:
> 
> Hi,
> 
> Thanks for the analysis.  Yes, all variable-length fields are immediately preceded by their length.
> 
> Rgds
> 
> Damon
> 
>> On 29 Dec 2015, at 12:49, Jeremy Poulter <jeremy at bigjungle.net> wrote:
>> 
>> If I am reading it correctly you have the total packet length (including the body) as the first byte so I am OK with the body size immediately proceeding the body data and in general this should be the same format for all variable length fields (no checked this is the case in detail).
>> 
>> Just my 2p worth.
>> 
>> Jeremy
>> 
>> On 24 Dec 2015 10:27, "Damon Hart-Davis" <damon at opentrv.uk> wrote:
>> Hi,
>> 
>> http://www.earth.org.uk/OpenTRV/stds/network/20151203-DRAFT-SecureBasicFrame.txt
>> 
>> Matthew W suggests that parsing (etc) might be significantly easier if the body length field was moved up somewhere in front of the variable-length ID field (or if the ID length was fixed, but I’m pretty certain that I don’t want to take that path).
>> 
>> I’m loathe to separate the body length from the body, but I’m interested in other views on this.  Simplicity in parsing/generating is very valuable for the sorts of small systems that this protocol is aimed at.
>> 
>> Will Santa bring us a working protocol?  B^>
>> 
>> Rgds
>> 
>> Damon
>