[dev] Proposal: coresource api

2017-12-11 Thread Gregg Reynolds
Hi list, I've got something working in OpenOCF that might fit with Iotivity. The idea is that, at least for debugging, we can database messages. On the client side we add a KEEP_MESSAGE flag to complement the KEEP_TRANSACTION flag. There is no reason a client should need to copy the response msg;

Re: [dev] which libcoap to use in master branch?

2017-12-11 Thread Gregg Reynolds
Hi Martin! On Dec 11, 2017 5:20 AM, "Martin Roesch" wrote: ... What scope do you have in mind? Minimal OCF support. Iotivity is chock-full of add-ons. They're fine, but I would prefer that add-ons be separate from the core engine. IOW I would like to have a minimal OCF kernel that devs can

Re: [dev] About SECURED mode

2017-12-11 Thread Philippe Coval
On 11/12/17 17:10, Mats Wichmann wrote: > On 12/11/2017 02:04 AM, Sun Lifeng 孙立峰 (690) wrote: >> Hi, dev, >> >> For IoTivity project, about SECURED mode, for my understanding, I should >> generate dat file for the SVR database. Am I right? yes you need to write ACL in json and then compile it c

Re: [dev] About SECURED mode

2017-12-11 Thread Mats Wichmann
On 12/11/2017 02:04 AM, Sun Lifeng 孙立峰 (690) wrote: > Hi, dev, > > For IoTivity project, about SECURED mode, for my understanding, I should > generate dat file for the SVR database. Am I right? > > If I do, how to generate that dat file? Is there a specific tool? Yes, json2cbor. it builds out

[dev] About SECURED mode

2017-12-11 Thread 690
Hi, dev, For IoTivity project, about SECURED mode, for my understanding, I should generate dat file for the SVR database. Am I right? If I do, how to generate that dat file? Is there a specific tool? Thanks Jason Sun ___ iotivity-dev mailing list iot

Re: [dev] which libcoap to use in master branch?

2017-12-11 Thread Martin Roesch
Hi > On Dec 5, 2017 1:25 PM, "Dave Thaler via iotivity-dev" > wrote: > ... >   Iotivity should really be rewritten to be a clean layer on top of the >normal libcoap APIs. > > +1000 I agree that IoTivity should use the normal libcoap. > Not just for coap, but because the code is pretty horribl