On Monday, 17 April 2017 23:45:48 PDT ??? (Uze Choi) wrote: > I believe, IoTivity should be extended with cloud further and the other > vertical such as industrial domain.
As replied to Dwarka, agreed and I think we actually need more. We need Fog, we need more verticals and we need the opposite of Cloud: the edge devices too. > For the better complete solution provider, Lightweight kernel is also > required. Sorry, what do you mean by lightweight kernel? Do you mean the IoTivity Constrained implementation? Or do you mean IoTivity should work on an operating system kernel? If it's the former, I agree and we have that. If it's the latter, I disagree vehemently. IoTivity should consume the product of other Open Source projects creating lightweight kernels. > The other is the organizational misalignment with OCF. > > IoTivity is currently separate open source project independent from OCF. > > However, relationship is not so good. Very few support from OCF (just infra > budget support) and from the OCF view it show very few response from OCF > request. To allow more power with more responsibility into each IoTivity > position will facilitate better communication. As I've explained in previous meetings and emails, I don't have a solution for the communication problem and the lack of attention from OCF and the majority of its member companies. I'll defer to you and to others who have more experience. > So that TSC organization looks good, but BSC proposal looks little bit > overhead considering OCF OSWG. Hmm Here's an interesting thing: as I've been designing this proposal for IoTivity, I was also designing an almost identical one for another open source project that is the reference implementation of the interoperability specifications produced by another industry group (sounds familiar?). I just took the IoTivity one and had them review it. It turns out it was a lot of overhead on the BSC / GB side. So the proposal for them is to have only a TSC. So I agree on scaling back the GB side for IoTivity. We don't have a budget that we need to manage, for example. However, beause IoTivity is a Linux Foundation Collaborative project, the GB needs to exist, even if they only meet twice a year just to check if everything is still going on correctly. That also leaves the Events and Community subcommittee. Events is definitely a non-technical task, but since we don't have money to sponsor anything anyway, do we need the work? If it's just the Developer Community that is left, we can justify it as a technical work. -- Thiago Macieira - thiago.macieira (AT) intel.com Software Architect - Intel Open Source Technology Center
