After lambda_solid is removed publicly, we want to keep it around as an internal API? Why?
Today
Today
sandyryza added a comment to D4868: RFC: deprecate imports of lambda_solid from top level.
dgibson added a comment to D4972: use instance_for_test in test_event_callback function.
We can find the real underlying issue using https://dagster.phacility.com/D4975, the migration error is masking the true cause
schrockn added a comment to D4972: use instance_for_test in test_event_callback function.
got it. I reopen-ed https://github.com/dagster-io/dagster/issues/3169 for tracking if you want to put failures somewhere
dgibson added a comment to D4972: use instance_for_test in test_event_callback function.
ok I no longer think that was it, I'm seeing this in other tests that use instance_for_test :) there's something funky going on with the event log.
Reform docs frontpage
sandyryza added a comment to D4749: Reform docs frontpage.
Ok, I'm going to merge this because I think it's still an improvement over our current state, even if there's opportunity for visual improvement.
Harbormaster failed remote builds in B20524: Diff 24885 for D4911: Differentiate between ModuleCodePoitner and PackageCodePointer on the gRPC server!
dgibson updated the diff for D4916: Add a RepositoryLocationOrigin class that separates out the loading instructions for a repo location from the actual repository data that it loads.
lint
dgibson updated the diff for D4911: Differentiate between ModuleCodePoitner and PackageCodePointer on the gRPC server.
remove mention of removing module in 0.10.0 since i'm not positive that's actually the plan
Upgrade Apollo
dish committed R1:4c4699076e22: [dagit] Empty definition state for empty pipelines (authored by dish).
[dagit] Empty definition state for empty pipelines
dish added a comment to D4931: Upgrade Apollo.
Should note that we are removing tests for rendering the type page, solids page, and playground. But let's burn it down. Curious to see what testing structure you have in mind...
bob updated the diff for D4889: Uses run tags instead of external pipeline tags for user defined k8s config..
- Prunes buildkite steps to only include integration tests.
- Adds dependency injection for k8s client on dagster-k8s run launcher.
- Fix the pruned buildkite pipeline. 0
- Fixes pruned buildkite pipeline. 1
dgibson updated the diff for D4916: Add a RepositoryLocationOrigin class that separates out the loading instructions for a repo location from the actual repository data that it loads.
comments and rebase
dpeng817 <cdguitar817@gmail.com> committed R1:da9b8de14434: Test for python set's dagstertypeloader (authored by dpeng817 <cdguitar817@gmail.com>).
Test for python set's dagstertypeloader
dgibson updated the diff for D4911: Differentiate between ModuleCodePoitner and PackageCodePointer on the gRPC server.
switch to --package-name
prha requested changes to D4941: Split RepositoryOrigin/PipelineOrigin hierarchy into ExternalOrigins and PythonOrigins.
This makes a lot of sense to me.
dgibson requested changes to D4915: [run-queue-5] Default runs coordinator (sends to external process).
other than the name this seems good to me, see inline. We can also do a single diff at the end of this stack where we do a big codemod to make all the final naming decisions, since we're going back and forth so much and renaming through a stack is a pain
we should take out the 's' otherwise this looks good :)
dgibson added a comment to D4891: [Run-queue-3] LaunchImmediateRunsCoordinator.
other than the naming stuff this is looking good to me, and we can do a rename pass at the end of the stack too. @alangenfeld did you have any blockers left here?
[dagit] Flip UTC offset sign in timezone picker
[dagit] Use Table for /scheduler
dgibson added inline comments to D4973: Fix test_sync_list_container_grpc running locally outside of BK.
max requested review of D4945: Load from module when possible in reconstructable.
max requested review of D4868: RFC: deprecate imports of lambda_solid from top level.