It is becoming apparent we need to mainatain our superset fork.
Let's backport to this fork:
- fixes done to support 3.5
- fixes for filter box javascript issues if any
It is becoming apparent we need to mainatain our superset fork.
Let's backport to this fork:
Status | Subtype | Assigned | Task | ||
---|---|---|---|---|---|
Resolved | odimitrijevic | T211706 Superset Updates | |||
Resolved | • fdans | T210687 Bug: can't make a YoY time series chart in Superset | |||
Resolved | elukey | T213488 Superset's rolling average feature results in error message | |||
Declined | • fdans | T211605 Upgrade Superset to 0.28.1 | |||
Declined | elukey | T211932 Port our "fixes" for python 3.5 to our superset fork https://github.com/wikimedia/incubator-superset |
With https://gerrit.wikimedia.org/r/#/c/operations/puppet/+/480041/ the SRE enabled us to use Python 3.6 on Stretch (without waiting for Buster), so in theory we could avoid using our fork and "wait" for stable releases from upstream.
While that would be ideal seeing the churn of changes in project and the low level of quality of master for 0.28 release I think we are probably going to need our fork going forward.
I'll create a staging environment next quarter and then hopefully 0.29 will be a bit more stable, if not we can keep the fork indeed. Would this work?
I'll create a staging environment next quarter and then hopefully 0.29 will be a bit more stable, if not we can keep the fork indeed. Would this work?
yes, but in the meantime we need to try fix T210687 which (correct me if i am wrong) requires us to backport to those changes to our fork (cc @fdans )
I am closing this task as we would not need to port fixes to 3.5 since we will be able to run in 3.6. It is likely we need to "patch" our fork for other reasons that have little to do with python versions.