Socket work

[mitch-gilmore]

This PR separates the ML and Socket work into a separate PRs to make reviews faster.

[jonathanm-12]

Code looks fine as it uses a simple TCP socket technique. The server_socket.py file serves as an effective intermediary between the three Python programs in our project.

[mitch-gilmore]

@dayannat @jonathanm-12 Is the server_socket.py file needed or just written for the latest demo?

[jonathanm-12]

@dayannat @jonathanm-12 Is the server_socket.py file needed or just written for the latest demo?

it will be needed

@mgilmore42

[mitch-gilmore]

@jonathanm-12

I guess I'm not understanding its utility. Can you please expand? My impression was the clients/servers would need to be embedded into the 3 parts (boardStreamer, ML, Virtual Environment). Like so..

boardStreamer ----(TCP)---> ML ----(TCP)---> Virtual Environment

Where boardStreamer serves to ML. ML is a client of boardStreamer and a server of Virtual Environment. And Virtual Environment is purely a client of ML. Similar to what was added to boardStreamer/__main__.py around line 100.

1. I this intuition incorrect?
2. If this it true how does an additional script fit into the pipeline?

To me it looks like just a demo to capture the data and acts purely as a client (against its name). Please explain where my understanding of its utility diverges from your understanding or how my understanding can be reconciled with your own.

Cancel

[mitch-gilmore]

@jonathanm-12 I removed the file for now until it utility can be better explained/defended. I also changed the port assignment to use the command-line argument instead of being hard coded.

[jonathanm-12]

Will approve once I see that this can work in sync with ML and virtual environment.

[dayannat]

@jonathanm-12

I guess I'm not understanding its utility. Can you please expand? My impression was the clients/servers would need to be embedded into the 3 parts (boardStreamer, ML, Virtual Environment). Like so..

boardStreamer ----(TCP)---> ML ----(TCP)---> Virtual Environment

Where boardStreamer serves to ML. ML is a client of boardStreamer and a server of Virtual Environment. And Virtual Environment is purely a client of ML. Similar to what was added to boardStreamer/__main__.py around line 100.

1. I this intuition incorrect?
2. If this it true how does an additional script fit into the pipeline?

To me it looks like just a demo to capture the data and acts purely as a client (against its name). Please explain where my understanding of its utility diverges from your understanding or how my understanding can be reconciled with your own.

This was meant to be a demo for the final prototype testing. As there was no ML model existing at the time, there could be no closed connection. Therefore, the purpose was to prove functionality of the socket programming. From here, what's left is to add sockets through the ML model that will receive data from the board and output the predictive results to the virtual environment.