Replies: 2 comments 2 replies
-
@captainsafia Thoughts? |
Beta Was this translation helpful? Give feedback.
-
This looks like a sensible breakdown for the API.
I have some hesitancies about this. In the past, we've gotten in trouble with implementation just the API for a feature without implementing the UI. For one, it means that we don't have a way to verify the behavior of something end-to-end and it inevitably brings up requests for the UI at some point. I certainly don't want my perspective to be the hold up here but I think it's one of those things where it is best to go all the way with it. @miduncan Based on your experience, do you think building out the UI is a large investment or is most of the hard work in figuring out the APIs and integration with the kernel? |
Beta Was this translation helpful? Give feedback.
-
Hey all!
Currently my team has a feature called “Variable Explorer”, which is a table that shows the current values of the variables in the kernel. I’ve attached a gif at the bottom to see it in action.
We’ve written this so that we can contribute the logic back to nteract (perhaps in its own package?). Essentially what would be added is
o Array of variables (name, type, length, value)
o isLoading (used for loading indicators when querying kernel)
o isEnabled (so it can be disabled for performance reasons when not shown)
o Provider used for querying kernel (only supports ipython currently)
o Set for redux state fields above
o RefreshVariables to re-query kernel
o Get for redux state fields above
o After a RUN_CELL, enableVariableExplorer, or refreshVariables action, the variable values will be refreshed and stored variables array in the redux state
Unfortunately, we don’t have the time allocated for creating a UI for this; however, since the variable info would all be accessible via selectors, this should be easy to hook-up for future UI efforts. Would this be something that nteract is interested in?
Thanks,
Michael
Beta Was this translation helpful? Give feedback.
All reactions