Archived Forum

I've made a start on a documentation page that describes a couple of the key parts of configuring emonhub.conf including the node decoding.

The first draft is available here:
https://github.com/openenergymonitor/emonhub/blob/emon-pi/configuration.md

I've also added the node decoder config to the top of the standard emontx, emonth and emonpi firmwares i.e: https://github.com/openenergymonitor/emonpi/blob/master/Atmega328/emonPi...

I've been discussing with Paul Burnell whether we should ship the emonpi with the default set of node decoders included as we currently do. Or would it be better to require the adding of node decoders manually? Running emonpi update also adds new node configuration automatically. Which may not be the ideal for users who are running more custom setups, discussion with Kevin here is a good example: http://openenergymonitor.org/emon/node/12081

All OEM hardware currently send packets that are an even number of bytes long, mostly consisting of 2 byte integers but also more recently 4 byte unsigned longs. The longs are always at the end of the packet and so it is possible to use emonhub.conf with no node decoders specified and data will appear in the nodes module or inputs module with values decoded correctly, apart from the unsigned long.

Perhaps it would be possible to add a checkbox in emonpi update to select either option and maybe even have the ability to specify in the emoncms config module whether to auto populate the nodes config on a new system.

Thoughts on this and the new config documentation would be very welcome.

I should note the documentation is for the emon-pi variant of emonhub.