Add workflow to run pyhubblenetwork CI tests - #8
Conversation
hunterhubble
commented
Jun 11, 2026
- On PRs to hubble-satnet-decoder, run sdr-docker, which will build with a local copy of hubble-satnet-decoder (not pip install), and then runs the pyhubblenetwork CI tests
|
Coincides with PRs from pyhubblenetwork and sdr-docker |
3938840 to
a0e6c38
Compare
|
I'm thinking do we need this? If we think about it, it seems circular in a sense that a PR to either 1 of these 3 (sdr-docker, satnet decoder, pyhubblenetwork) will trigger a rebuild. It doesn't neccessary means that it'll pass. The layer of dependencies are: pyhubblenetwork -> depends on -> sdr-docker -> depends on -> this repo (satnet decoder). For example, if there is a protocol change, the upper layer is definitely going to fail. |
I agree it's not a thorough test, it is not meant to test the decoder, maybe I can look into how to validate the decoder against another source of truth. This is more of a smoke test to ensure that the changes to hubble-satnet-decoder still lets us build sdr-docker and use the pyhubblenetwork tools as we normally could. |
|
@hunterhubble I re-ran CI because the PR that you added for running pyhubblenetwork tests just got merged to |
Thank you for looking, I'll try and see why that is |
- On PRs to hubble-satnet-decoder, run sdr-docker, which will build with a local copy of hubble-satnet-decoder (not pip install), and then runs the pyhubblenetwork CI tests Signed-off-by: hunterhubble <hunter@hubble.com>
a0e6c38 to
b5cc211
Compare
|
@HongNguyen635 found the issue, the way I copied the /decoder-src directory to sdr-docker so that it could build with the local version of hubble-satnet-decoder instead of the pip was not compatible with the gh action to build the container. This new way does that correctly and it just passed. Thank you for flagging |
HongNguyen635
left a comment
There was a problem hiding this comment.
Looking good. Your explanation about why you added this workflow makes sense to me.