WIP - [WFM Wrapper] Move device specific methods into device lib & invert dependencies between devices and wfm #3
Conversation
twoshark
changed the title
lib/Relay/Relay.cpp
Outdated
|
|
||
|
|
||
|
|
||
| void Relay::SendHTTP(WiFiClient client) |
There was a problem hiding this comment.
Do you think we should keep this? Weve seen that using a UI kinda wrecks the webserver.
There was a problem hiding this comment.
Theoretically, that is improved by the division of work across cores. It's nice to have it as a debug option for individual devices if the overhead ends up not being an issue.
Will test!
| Relay relay = Relay(); | ||
| TaskHandle_t WebServerTaskHandle; | ||
| TaskHandle_t IOTaskHandle; | ||
| WiFiManagerWrapper wfm = WiFiManagerWrapper(); |
There was a problem hiding this comment.
since this WiFiManagerWrapper (we could prolly change the name of this at some point soon) still handles the /specification endpoint. I think we will want to inject whatever device information we want to include in that spec response. The most basic example could be the device_type. Which in this case would be relay?
The benefit of having the wfm do it for all devices will be the standardization of the spec responses.
There was a problem hiding this comment.
Hmm this is a tricky one, I actually intended to move this path into the Relay class. I just hadn't done it yet.
We can inject the spec in when we instantiate a new WFM in Relay, but I think it should be a shared type that contains maps of paths/parameters/etc. Then wfmw can validate the input and pull data from known keys.
That way the wfm wrapper can stay generic and there's closure around device business logic in the lib/Relay/
- moved shouldSaveConfig into Wfm Wrapper Class - updated shouldSaveConfig to use an anon func that sets shouldSaveConfig - delete unused mutex
No description provided.