-
Notifications
You must be signed in to change notification settings - Fork 1.3k
Open
Description
I wonder if anyone has intuition or empirical data on performance differences across website technologies. In our internal evals we created a plain JavaScript site to test our wrappers and we consistently get near 100 percent accuracy. But we see some Angular sites underperform by a very wide margin.
- Simple input field interactions using the observe then act pattern fail more than we would expect.
- Dropdowns that render elements at the root of the document such as hover style dropdowns succeed only about ten percent of the time
I am about to create a dedicated Angular eval set, but I am curious whether others have seen similar issues across different frameworks. I can understand how this might happen if models are trained more heavily on React than Angular, but I am mostly looking for solutions or workarounds.
Metadata
Metadata
Assignees
Labels
No labels