Closed dirkmueller closed 1 month ago
Please don't. Instead use
osc = Osc(public=True)
instead so that we can rely on the error handling code inside py_obs
why do we need this error handling for this function? it just tries to inject authentication which is not necessary here, and then retries with fallback on error code 500. I don't find that sensible behavior for updating the versions. If there's error 5xx there is an overload and just blindly retrying is not useful imho for this functionality.
also btw the current eventloop "loop/retrying" logic is failing in CI so the package_verisons are outdated, which is why I started to look into it.
Created a staging project on OBS for Tumbleweed: home:defolos:BCI:Staging:Tumbleweed:Tumbleweed-1482
Changes pushed to branch Tumbleweed-1482
as commit a1f64eaf24083f5a122ba3c0f52bdd1f64355b16
Build succeeded ✅
Build succeeded ✅
To run BCI-tests against this PR, use the following command:
OS_VERSION=tumbleweed TARGET=custom BASEURL=registry.opensuse.org/home/defolos/bci/staging/tumbleweed/tumbleweed-1482/ tox -- -n auto
obsolete, was changed via #1500
This requires setting OSC_USER and OSC_PASSWORD which is just annoying and unnecessary. we can do this with a single interconnect api call just as well.