Camera disconnected + stream normalization improvements #1701
Conversation
There was a problem hiding this comment.
What is the source and license for this image?
There was a problem hiding this comment.
Didn't think of that when I found it, but after looking back it seems it was made by someone on a GitHub PR in 2020:
FAForever/downlords-faf-client#1665
Unsure if this is an issue. Opinions on this? There weren't many other good ones I could find that looked clean in this.
There was a problem hiding this comment.
The repo is MIT licensed. But does that cover non-merged content? I’m not sure.
|
If I try to "start calibration", I don't see the processed stream (it almost seems to freeze). But if I manually navigate to http://10.17.36.12:1181/stream.mjpg and http://10.17.36.12:1182/stream.mjpg , I do. And if I reload the webpage, I see it too. |
Odd, I can't seem to reproduce this, is it device/camera specific? Quickly tested on vite dev run with arducam 9281 |
Images are before and after comparison.
Does the following:
Fixes logic for disconnected cameras in camera matching view to properly indicate the actual status of the camera. Previously disconnected cameras were falsely flagged as mismatches.
Sorts camera modules by (active, connected) / mismatch -> (active, disconnected) -> deactivated -> unassigned
Massively (as far as I can tell) improves camera stream height normalization logic in camera matching view, and guts all the unnecessary logic previously used to accomplish this
(verification that it works, for me at least)
Disclaimer: if anyone was super attached to the yellow loading spinner, it had to go to make this cleaner stream normalization work out. It's one of these guys now, which I happen to think is cleaner looking anyway:
P.S. camera lost image, don't think I didn't see you...