this is a formatted list of some of my text based remarks…
- Inconsistent Data Requests:
- Variable number of individual data requests across similar tests (e.g., 3, 4, or 5 requests) despite identical conditions.
- Unexpected increase in data requests when conditions are placed on actions versus event triggers.
- Variable Work Unit (WU) Costs:
- Creating a “Broker” consistently incurs higher WU costs (2.12 WU) compared to other “create new thing” actions (0.62 WU).
- Discrepancies in WU charges for similar actions, such as “create a new blog post” costing 1.43 WU versus expected totals.
- Unexpected Conditional Evaluation Charges:
- Some tests show unexpected charges for condition evaluations (e.g., 0.32 WU) while identical conditions in other tests do not incur these charges.
- Inconsistent application of conditional evaluation charges, leading to confusion over when and why they are applied.
- Search Charge Anomalies:
- Expected search charges (e.g., 0.3 WU) were missing in certain tests, especially when users were not logged in.
- Conditions requiring searches do not consistently trigger the expected data fetch charges.
- Workflow vs. Data Fetching Charges:
- Charges are sometimes labeled as “workflow” instead of specific actions like data fetching or conditional evaluations, causing ambiguity.
- Inconsistent labeling leads to difficulty in tracking and understanding charge sources.
- Ghost Bar Scenarios:
- Certain actions (e.g., at timestamps 1:54, 1:56, 4:52) show no associated charges, indicating that conditions or searches may not have been performed.
- Differences Based on User Authentication:
- Logged-out users trigger different charge patterns, such as workflow charges combined with individual data request charges, compared to logged-in users.
- Authentication status affects how conditions and actions are evaluated and billed.
- Charge Calculation Discrepancies:
- Total WU charges for actions (e.g., 1.12 WU for create) do not align with expected sums based on individual component charges (e.g., 0.62 WU for create + 0.3 WU for search = 0.92 WU).
- Inconsistent charge totals across similar actions make it difficult to predict billing accurately.
- Inconsistent Labeling of Charges:
- Charges related to conditional evaluations and data fetching are labeled inconsistently, sometimes under “workflow” instead of their specific action types.
- This inconsistency complicates the identification and understanding of charge sources.
- Conditions on Event Triggers:
- Placing conditions on event triggers results in different and often unexpected behavior compared to placing conditions directly on actions.
- Server-side evaluations on event triggers do not consistently align with expected charge patterns.
- Workflow Charge Variations Across Runs:
- Subsequent runs of similar workflows exhibit different charge applications, such as the first run having only a workflow charge and the second run adding data request charges.
- Variability between workflow executions complicates charge predictability and consistency.
I created these tests as I put up a post on conditions and security and some other users were adamant there are no issues and everything works as expected. Unfortunately, testing yields pretty much exactly the opposite.