Re: [LEDE-DEV] Question about rpcd uci::state

2017-03-06 Thread Jo-Philipp Wich
Hi, my plan was to add a boolean option to the "get" procedure which - if set - will temporarely switch the savedir to an invalid path (like /dev/null), then switch it back to the previous value. The rpcd daemon currently switches the savedir per session/call anyway so it should be a minor chang

Re: [LEDE-DEV] Question about rpcd uci::state

2017-03-06 Thread Zefir Kurtisi
On 03/06/2017 01:15 PM, Jo-Philipp Wich wrote: > Hi, > >> the uci ubus service (part of rpcd) provides the method 'state' which from >> reading >> the source code is expected to provide the current state of an option >> ignoring >> pending changes. > > No it is a method to provide access to the

Re: [LEDE-DEV] Question about rpcd uci::state

2017-03-06 Thread Jo-Philipp Wich
Hi, > the uci ubus service (part of rpcd) provides the method 'state' which from > reading > the source code is expected to provide the current state of an option ignoring > pending changes. No it is a method to provide access to the legacy concept of "state variables" which are epheramal uci va

[LEDE-DEV] Question about rpcd uci::state

2017-03-06 Thread Zefir Kurtisi
Hi Jo, all, the uci ubus service (part of rpcd) provides the method 'state' which from reading the source code is expected to provide the current state of an option ignoring pending changes. At my side it does not (shows the same as 'get'): > root@LEDE:~# ubus call uci get '{"config":"system", "