Open MichaelTiemannOSC opened 2 years ago
HumairAK
commented
2 years ago Don't think we want to start maintaining a separate image for superset that is not backed by a proper release. To ensure stability we should stick the the released versions, and wait for this to be included in either 1.4 or whichever version it makes it in.
@erikerlandson @rimolive
erikerlandson
commented
2 years ago If this is blocking something important and time sensitive I am OK with going off release, but I agree with Humair that it is preferable to stick to supported releases.
So the question is where this lives on the risk/reward curve.
MichaelTiemannOSC
commented
2 years ago If somebody can spin this up for me, I can test it and potentially vote:
erikerlandson
commented
2 years ago Would be fine with deploying a release candidate
MichaelTiemannOSC
commented
2 years ago n.b. The SuperSet team rescinded the vote for rc1 and have retreated to bugfix, cherry-pick other features, and offer a new rc2 for a vote. That was expected to happen in early December, but has not happened yet. We should delay upgrading until rc2 gets some good preliminary reviews.
MichaelTiemannOSC
commented
2 years ago rc3 is available for review and testing: https://github.com/apache/superset/releases/tag/1.4.0rc3
MichaelTiemannOSC
commented
2 years ago rc4 is now available for review and testing: https://dist.apache.org/repos/dist/dev/superset/1.4.0rc4/
You can vote for it here: https://github.com/apache/superset/discussions/18018
MichaelTiemannOSC
commented
2 years ago There was a small glitch in the 1.4.0 cherry-picking process that caused one feature to interact with another. The fix is trivial and is being suggested to be merged into 1.4.1: https://lists.apache.org/list.html?dev@superset.apache.org
But if we deploy 1.4.0, we should apply this one-liner so as not to lose one of the cherries picked.
MichaelTiemannOSC
commented
2 years ago And just like that, 1.4.0 has been released: https://github.com/apache/superset/releases/tag/1.4.0
HumairAK
commented
2 years ago same as: https://github.com/os-climate/os_c_data_commons/issues/116#issuecomment-1010013323 @rimolive or someone from odh will need to provide an odh superset image upgrade for us to leverage
MichaelTiemannOSC
commented
2 years ago Tracking here: https://github.com/opendatahub-io/odh-images/issues/11
Describe the bug The deck.gl visualization plugin renders legend text as NULL. There is a PR that fixes it (from back in August), but it was missed in the multirepo merge.
To Reproduce See https://github.com/apache-superset/superset-ui-plugins-deckgl/pull/43
Expected behavior Legends have proper text, not NaNs
Screenshots If applicable, add screenshots to help explain your problem.
Additional context https://github.com/apache/superset/issues/15738
Noise has been made that this feature should land in 1.4, but I don't know. We should merge this PR so that we can render more kinds of maps with legends. And hopefully it will be fixed when we do our next major version update.