Closed Dinsmoor closed 6 months ago
I agree that getting rid of that duplicate file is good. It appears that the changes to TCPHostInterface.c are only improved comments, with no functional changes made. I read through them, and it appears to be a good improvement. However, I haven't studied this portion of the code well enough yet to say that I can confirm that everything the new comments say is accurate.
I did notice that a few comments refer to the UDP based GUI, seeming to suggest that this is nothing but a waterfall display. While data for a waterfall display is sent to that GUI, additional data about the activity and state of ardop is also sent to the GUI where it is presented to the user. This combination of data can provide useful feedback to the user, especially when an ARQ connection is failing. That GUI also has the ability to send a few selected messages back to ardop requesting that IDFrames or 2-tone test signals be transmitted, or controlling what sort of data (Waterfall vs. Spectrum vs. neither) is sent to the GUI. For now, I think that we should continue to support this interface, though it may be phased out later.
I ran out of steam when I got to the UDP section of that file. It was confusing to figure out which comments left by previous authors were still relevant. For example, there is no two byte CRC on any TCP interface at all, other than what the socket itself provides.
I'm going to leave this file alone for now though, and start looking at improving documentation elsewhere.
Corrected documentation of TCP interfaces. Partially documented UDP interface - we may consider removing this code if we don't begin supporting ARDOP_GUI into this project. Removed unneeded file duplicate (the DESKTOP-MHE5LO8.c only had one line of code different, and was likely one used in testing by a previous author)