Page MenuHomeElementl
Feed Advanced Search

Yesterday

cdecarolis requested review of D8360: [docs] respin io manager testing section.
Mon, Jun 14, 9:53 PM
cdecarolis accepted D8356: don't require context arg when solid has required resources.

Yea this makes sense to me. The step launcher case already made me a bit uneasy.

Mon, Jun 14, 9:19 PM
cdecarolis added a comment to D8362: rename SensorExecutionContext to SensorContext.
In D8362#218480, @prha wrote:

+1 on moving way from ExecutionContext.

I think SensorContext is fine... did you also consider SensorEvaluationContext/ ScheduleEvaluationContext? I think we've been pretty good in the docs about "evaluating" sensors & schedules and "executing" pipelines.

Mon, Jun 14, 7:44 PM
cdecarolis requested review of D8362: rename SensorExecutionContext to SensorContext.
Mon, Jun 14, 7:11 PM
cdecarolis updated the diff for D8275: Make solid invocation directly call the underlying decorated fxn.

Up

Mon, Jun 14, 3:56 PM
cdecarolis updated the diff for D8275: Make solid invocation directly call the underlying decorated fxn.

up

Mon, Jun 14, 3:51 PM
cdecarolis updated the diff for D8275: Make solid invocation directly call the underlying decorated fxn.

Rearrange implementation around flag for decorators requiring context arguments, fix tests

Mon, Jun 14, 3:48 PM

Thu, Jun 10

cdecarolis updated the diff for D8275: Make solid invocation directly call the underlying decorated fxn.

Remove compute wrapper override, add DecoratedSolidFunction class to determine whether solid fxn came from decorator

Thu, Jun 10, 8:35 PM
cdecarolis updated the diff for D8275: Make solid invocation directly call the underlying decorated fxn.

Add solid compute wrapper override flag

Thu, Jun 10, 6:35 PM
cdecarolis updated the test plan for D8331: [docs] add resource testing section.
Thu, Jun 10, 5:39 PM
cdecarolis requested review of D8331: [docs] add resource testing section.
Thu, Jun 10, 5:08 PM
cdecarolis added inline comments to D8275: Make solid invocation directly call the underlying decorated fxn.
Thu, Jun 10, 4:23 PM
cdecarolis added inline comments to D8275: Make solid invocation directly call the underlying decorated fxn.
Thu, Jun 10, 3:54 PM
cdecarolis updated the diff for D8275: Make solid invocation directly call the underlying decorated fxn.

Up

Thu, Jun 10, 3:51 PM
cdecarolis updated the diff for D8275: Make solid invocation directly call the underlying decorated fxn.

Add output type checking and validation

Thu, Jun 10, 3:50 PM

Wed, Jun 9

cdecarolis accepted D8310: docs: add SolidExecutionContext to relevant APIs for Solids.

I'm on board

Wed, Jun 9, 4:36 PM
cdecarolis added a comment to D8275: Make solid invocation directly call the underlying decorated fxn.

I think yes, we are still validating inputs right?

Wed, Jun 9, 4:04 PM
cdecarolis requested review of D8306: Resolve Versions fix.
Wed, Jun 9, 3:37 PM
cdecarolis added a comment to D8275: Make solid invocation directly call the underlying decorated fxn.

What's the verdict on output validation? Is that something we want to pursue? cc @alangenfeld @sandyryza

Wed, Jun 9, 3:30 PM
cdecarolis added inline comments to D8275: Make solid invocation directly call the underlying decorated fxn.
Wed, Jun 9, 3:29 PM
cdecarolis requested review of D8291: Hook Invocation.
Wed, Jun 9, 3:01 PM

Tue, Jun 8

cdecarolis added a comment to D8239: Sensor invocation, remove evaluate_tick from API docs, change testing story.

Do we want to make the context arg itself optional when defining the sensor function?

Tue, Jun 8, 4:46 PM
cdecarolis added a comment to D8275: Make solid invocation directly call the underlying decorated fxn.

its a bit complex but we *could* do output validation if we wanted

do we want to?

Tue, Jun 8, 2:20 PM

Mon, Jun 7

cdecarolis requested review of D8277: Small bugfix for logger invocation, and add testing section to logger docs.
Mon, Jun 7, 10:41 PM
cdecarolis requested review of D8275: Make solid invocation directly call the underlying decorated fxn.
Mon, Jun 7, 8:44 PM

Fri, Jun 4

cdecarolis updated the diff for D7999: build_init_resource_context.

Making resource decorate a cm fixes lint issue

Fri, Jun 4, 4:01 PM
cdecarolis updated the test plan for D8239: Sensor invocation, remove evaluate_tick from API docs, change testing story.
Fri, Jun 4, 3:50 PM
cdecarolis updated the diff for D8239: Sensor invocation, remove evaluate_tick from API docs, change testing story.

Update docs to reflect optionality of context arg on sensor invocation

Fri, Jun 4, 3:25 PM

Thu, Jun 3

cdecarolis updated the diff for D8184: Make pending node invocations pulled from solids invokable.

Up

Thu, Jun 3, 9:52 PM
cdecarolis updated the diff for D8184: Make pending node invocations pulled from solids invokable.

Re-add assert_in_composition stuff

Thu, Jun 3, 9:49 PM
cdecarolis requested review of D8239: Sensor invocation, remove evaluate_tick from API docs, change testing story.
Thu, Jun 3, 9:29 PM
cdecarolis added inline comments to D7999: build_init_resource_context.
Thu, Jun 3, 9:16 PM
cdecarolis requested review of D8184: Make pending node invocations pulled from solids invokable.

bump

Thu, Jun 3, 8:21 PM
cdecarolis planned changes to D8184: Make pending node invocations pulled from solids invokable.
Thu, Jun 3, 8:20 PM
cdecarolis added inline comments to D7999: build_init_resource_context.
Thu, Jun 3, 7:22 PM
cdecarolis updated the diff for D7999: build_init_resource_context.

up

Thu, Jun 3, 5:44 PM
cdecarolis updated the diff for D7999: build_init_resource_context.

Up

Thu, Jun 3, 4:15 PM
cdecarolis updated the diff for D7999: build_init_resource_context.

respun using new invocation strategy, renamed context builder, fixed config issues

Thu, Jun 3, 4:03 PM

Wed, Jun 2

cdecarolis updated the diff for D8140: initialize method for logger that enables default config.

Updated to use invocation instead of initialize

Wed, Jun 2, 7:20 PM
cdecarolis updated the diff for D8184: Make pending node invocations pulled from solids invokable.

Update comment block

Wed, Jun 2, 6:17 PM
cdecarolis updated the diff for D8184: Make pending node invocations pulled from solids invokable.

Explain aliased error message tests

Wed, Jun 2, 5:57 PM
cdecarolis updated the diff for D8184: Make pending node invocations pulled from solids invokable.

Save along attributes from invocations

Wed, Jun 2, 5:52 PM
cdecarolis updated the diff for D8168: [docs] testing partition sets, and putting partition set examples under test.

Ignore isort for testing docs example so that we can include import

Wed, Jun 2, 3:13 PM
cdecarolis updated the diff for D8168: [docs] testing partition sets, and putting partition set examples under test.

Up

Wed, Jun 2, 1:04 PM

Tue, Jun 1

cdecarolis requested review of D8189: Add decorated fxn properties to schedule and sensor.
Tue, Jun 1, 10:21 PM
cdecarolis updated the diff for D8168: [docs] testing partition sets, and putting partition set examples under test.

Make section heading experimental, include validate_run_config import

Tue, Jun 1, 8:30 PM
cdecarolis added inline comments to D8168: [docs] testing partition sets, and putting partition set examples under test.
Tue, Jun 1, 7:23 PM
cdecarolis updated the summary of D8184: Make pending node invocations pulled from solids invokable.
Tue, Jun 1, 6:26 PM
cdecarolis planned changes to D8140: initialize method for logger that enables default config.

Respin with call

Tue, Jun 1, 6:25 PM
cdecarolis added inline comments to D8155: Reorg implementation of config eval on solid invocation.
Tue, Jun 1, 6:19 PM
cdecarolis requested review of D8184: Make pending node invocations pulled from solids invokable.
Tue, Jun 1, 5:59 PM

Fri, May 28

cdecarolis requested review of D8168: [docs] testing partition sets, and putting partition set examples under test.
Fri, May 28, 9:38 PM
cdecarolis requested review of D8163: [memoization] ensure that versions are properly passed from configured.
Fri, May 28, 8:55 PM
cdecarolis requested review of D8155: Reorg implementation of config eval on solid invocation.
Fri, May 28, 3:15 PM

Thu, May 27

cdecarolis added inline comments to D8140: initialize method for logger that enables default config.
Thu, May 27, 10:03 PM
cdecarolis added inline comments to D8140: initialize method for logger that enables default config.
Thu, May 27, 10:02 PM
cdecarolis published D8147: Revert tutorial changes for solid invocation, make testing docs around solid invocation experimental for review.
Thu, May 27, 7:30 PM
cdecarolis updated the diff for D8133: Enable default config values with direct solid invocation.

Added more complex test cases, fixed docstring of BoundSolidExecutionContext, remove unnecessary if block

Thu, May 27, 6:50 PM
cdecarolis requested review of D8143: [easy] fix docstring on sensor execution context, make test names more informative.
Thu, May 27, 5:35 PM
cdecarolis updated the diff for D8140: initialize method for logger that enables default config.

add UnboundInitLoggerContext, which gets bound into InitLoggerContext during initialization

Thu, May 27, 5:06 PM
cdecarolis requested review of D8140: initialize method for logger that enables default config.
Thu, May 27, 5:03 PM
cdecarolis updated the diff for D8054: [docs] Schedule testing section, put schedule examples under test.

Up

Thu, May 27, 4:38 PM
cdecarolis updated the diff for D8135: Make instance_ref optional on schedule context, and instance optional on build_schedule_context.

Up after pins

Thu, May 27, 4:36 PM
cdecarolis published D8135: Make instance_ref optional on schedule context, and instance optional on build_schedule_context for review.

unrelated pin errors

Thu, May 27, 3:30 PM
cdecarolis updated the diff for D8054: [docs] Schedule testing section, put schedule examples under test.

Up

Thu, May 27, 3:30 PM
cdecarolis planned changes to D7999: build_init_resource_context.

align name with InitResourceContext, spin through some tests on internal, fix default config

Thu, May 27, 3:20 PM
cdecarolis updated the diff for D8054: [docs] Schedule testing section, put schedule examples under test.

Update with instance-optional changes

Thu, May 27, 3:17 PM
cdecarolis updated the test plan for D8054: [docs] Schedule testing section, put schedule examples under test.
Thu, May 27, 3:17 PM
cdecarolis updated the summary of D7999: build_init_resource_context.
Thu, May 27, 12:19 AM
cdecarolis accepted D8099: Add the mode's available resources to error message if resource requirements check fails.

Awesome. Thanks for tackling this!

Thu, May 27, 12:17 AM

Wed, May 26

cdecarolis requested review of D8133: Enable default config values with direct solid invocation.
Wed, May 26, 11:55 PM
cdecarolis added a comment to D7999: build_init_resource_context.

Will add a follow up diff to this to migrate over some demo tests

Wed, May 26, 7:53 PM
cdecarolis updated the diff for D8058: [docs] Expand section on testing sensors.

get_execution_data -> evaluate_tick

Wed, May 26, 7:24 PM
cdecarolis added a comment to D7999: build_init_resource_context.

Will add a follow up diff to this to migrate over some demo tests

Wed, May 26, 6:53 PM
cdecarolis added inline comments to D7999: build_init_resource_context.
Wed, May 26, 6:50 PM
cdecarolis added inline comments to D7999: build_init_resource_context.
Wed, May 26, 6:49 PM
cdecarolis requested changes to D8099: Add the mode's available resources to error message if resource requirements check fails.

either re-orienting language around providing resource for resource key / switching resource key, or some sort of levenshtein distance "did you mean __?" error message (not sure if we have a precedent for something like this in the codebase yet)

Wed, May 26, 4:49 PM
cdecarolis added a comment to D8099: Add the mode's available resources to error message if resource requirements check fails.

In cases where we're reasonably sure someone made a typo; i think it makes sense to communicate that there's a resource key that is similar to the one they provided. In the general case of someone missing a resource however, I'm not sure what the benefit of displaying keys is, however. Would the intention be that someone could see which resource keys are available, and change their definition to use that resource key instead? If that's the case, I think we might want to orient the language more around resource keys than definitions, since that's what we would expect people to be switching around.

Wed, May 26, 4:47 PM
cdecarolis updated the diff for D8112: get_execution_data -> evaluate_tick.

Get rid of weird vscode artifacts

Wed, May 26, 3:53 PM
cdecarolis added inline comments to D8058: [docs] Expand section on testing sensors.
Wed, May 26, 3:44 PM
cdecarolis updated the diff for D8112: get_execution_data -> evaluate_tick.

evaluate_tick

Wed, May 26, 3:43 PM
cdecarolis retitled D8112: get_execution_data -> evaluate_tick from get_execution_data -> do_tick to get_execution_data -> evaluate_tick.
Wed, May 26, 3:41 PM
cdecarolis planned changes to D8112: get_execution_data -> evaluate_tick.

Am on board with evaluate_tick @prha

Wed, May 26, 3:27 PM
cdecarolis updated the diff for D8053: Enable resource dependencies within memoizable io managers.

Address comments: get rid of defaults on get_output_context, pass resources directly to get_output_context

Wed, May 26, 3:26 PM
cdecarolis requested review of D8112: get_execution_data -> evaluate_tick.
Wed, May 26, 3:00 PM
cdecarolis updated the diff for D8079: build_init_logger_context.

Up

Wed, May 26, 2:53 AM
cdecarolis updated the diff for D8079: build_init_logger_context.

Update name to match context, make definition optional.

Wed, May 26, 2:46 AM
cdecarolis added inline comments to D8079: build_init_logger_context.
Wed, May 26, 1:55 AM
cdecarolis updated the diff for D8097: Update IO manager tests to use build_i/o_contexts.

Up

Wed, May 26, 1:54 AM

Tue, May 25

cdecarolis requested review of D8097: Update IO manager tests to use build_i/o_contexts.

Unrelated mypy error

Tue, May 25, 10:04 PM
cdecarolis added a reviewer for D8058: [docs] Expand section on testing sensors: prha.
Tue, May 25, 7:36 PM
cdecarolis updated the test plan for D8058: [docs] Expand section on testing sensors.
Tue, May 25, 7:35 PM
cdecarolis updated the summary of D8079: build_init_logger_context.
Tue, May 25, 5:57 PM
cdecarolis updated the diff for D8058: [docs] Expand section on testing sensors.

Add get_execution_data to api ref, update overview to reflect this

Tue, May 25, 5:54 PM
cdecarolis requested review of D8082: Make instance_ref optional on SensorExecutionContext to enable build_sensor_context() without instance.
Tue, May 25, 4:58 PM
cdecarolis requested review of D8079: build_init_logger_context.
Tue, May 25, 4:32 PM
cdecarolis accepted D8077: Update hello world code snippet in README.

Nice!

Tue, May 25, 3:53 PM
cdecarolis added inline comments to D8054: [docs] Schedule testing section, put schedule examples under test.
Tue, May 25, 3:43 PM
cdecarolis added inline comments to D8054: [docs] Schedule testing section, put schedule examples under test.
Tue, May 25, 3:42 PM
cdecarolis requested review of D8063: [mypy] logger definition.
Tue, May 25, 3:25 PM