2020.3.69010

2020.3.0 (19 March 2020)

Enhancements

1. Make interactive window wrap like the notebook editor does.
   (#4466)
2. Support scrolling beyond the last line in the notebook editor and the interactive window. Uses the editor.scrollBeyondLastLine setting.
   (#7892)
3. Allow user to override the arguments passed to Jupyter on startup. To change the arguments, run the 'Python: Specify Jupyter command line arguments" command.
   (#8698)
4. When entering remote Jupyter Server, default the input value to uri in clipboard.
   (#9163)
5. Added a command to allow users to select a kernel for a Notebook.
   (#9228)
6. When saving new notebooks, default to the current workspace folder.
   (#9331)
7. When the output of a cell gets trimmed for the first time, the user will be informed of it and which setting changes it.
   (#9401)
8. Change the parameters for when a Data Science survey prompt comes up. After opening 5 notebooks (ever) or running 100 cells (ever).
   (#10186)
9. Show quickfixes for launch.json.
   (#10245)

Fixes

1. Jupyter autocompletion will only show magic commands on empty lines, preventing them of appearing in functions.
   (#10023)
2. Remove extra lines at the end of the file when formatting with Black.
   (#1877)
3. Capitalize Activate.ps1 in code for PowerShell Core on Linux.
   (#2607)
4. Change interactive window to use the python interpreter associated with the file being run.
   (#3123)
5. Make line numbers in errors for the Interactive window match the original file and make them clickable for jumping back to an error location.
   (#6370)
6. Fix magic commands that return 'paged' output.
   (#6900)
7. Ensure model is updated with user changes after user types into the editor.
   (#8589)
8. Fix latex output from a code cell to render correctly.
   (#8742)
9. Toggling cell type from code to markdown will not set focus to the editor in cells of a Notebook.
   (#9102)
10. Remove whitespace from code before pushing to the interactive window.
    (#9116)
11. Have sys info show that we have connected to an existing server.
    (#9132)
12. Fix IPython.clear_output to behave like Jupyter.
    (#9174)
13. Jupyter output tab was not showing anything when connecting to a remote server.
    (#9177)
14. Fixed our css generation from custom color themes which caused the Data Viewer to not load.
    (#9242)
15. Allow a user to skip switching to a kernel if the kernel dies during startup.
    (#9250)
16. Clean up interative window styling and set focus to input box if clicking in the interactive window.
    (#9282)
17. Change icon spacing to match vscode icon spacing in native editor toolbars and interactive window toolbar.
    (#9283)
18. Display diff viewer for ipynb files without opening Notebooks.
    (#9395)
19. Python environments will not be activated in terminals hidden from the user.
    (#9503)
20. Disable Restart Kernel and Interrupt Kernel buttons when a kernel has not yet started.
    (#9731)
21. Fixed an issue with multiple latex formulas in the same '$$' block.
    (#9766)
22. Make notebook editor and interactive window honor undocumented editor.scrollbar.verticalScrollbarSize option + increase default to match vscode.
    (#9803)
23. Ensure that invalid kernels don't hang notebook startup or running.
    (#9845)
24. Switching kernels should disable the run/interrupt/restart buttons.
    (#9935)
25. Prompt to install pandas if not found when opening the Data Viewer.
    (#9944)
26. Prompt to reload VS Code when changing the Jupyter Server connection.
    (#9945)
27. Support opening spark dataframes in the data viewer.
    (#9959)
28. Make sure metadata in a cell survives execution.
    (#9997)
29. Fix run all cells to force each cell to finish before running the next one.
    (#10016)
30. Fix interrupts from always thinking a restart occurred.
    (#10050)
31. Do not delay activation of extension by waiting for terminal to get activated.
    (#10094)
32. LiveShare can prevent the jupyter server from starting if it crashes.
    (#10097)
33. Mark poetry.lock file as toml syntax.
    (thanks to remcohaszing)
    (#10111)
34. Hide input in Interactive Window based on the setting allowInput.
    (#10124)
35. Fix scrolling for output to consistently scroll even during execution.
    (#10137)
36. Correct image backgrounds for notebook editor.
    (#10154)
37. Fix empty variables to show an empty string in the Notebook/Interactive Window variable explorer.
    (#10204)
38. In addition to updating current working directory also add on our notebook file path to sys.path to match Jupyter.
    (#10227)
39. Ensure message (about trimmed output) displayed in an output cell looks like a link.
    (#10231)
40. Users can opt into or opt out of experiments in remote scenarios.
    (#10232)
41. Ensure to correctly return env variables of the activated interpreter, when dealing with non-workspace interpreters.
    (#10250)
42. Update kernel environments before each run to use the latest environment. Only do this for kernel specs created by the python extension.
    (#10255)
43. Don't start up and shutdown an extra Jupyter notebook on server startup.
    (#10311)
44. When you install missing dependencies for Jupyter successfully in an active interpreter also set that interpreter as the Jupyter selected interpreter.
    (#10359)
45. Ensure default host is not set, if connect or listen settings are available.
    (#10597)

Code Health

1. Use the new VS Code filesystem API as much as possible.
   (#6911)
2. Functional tests using real jupyter can take 30-90 seconds each. Most of this time is searching for interpreters. Cache the interpreter search.
   (#7997)
3. Use Python 3.8 in tests run on Azure DevOps.
   (#8298)
4. Display Commands related to Interactive Window and Notebooks only when necessary.
   (#8869)
5. Change cursor styles of buttons pointer in Interactive Window and Native Editor.
   (#9341)
6. Update Jedi to 0.16.0.
   (#9765)
7. Update version of VSCode in package.json to 1.42.
   (#10046)
8. Capture mimetypes of cell outputs.
   (#10182)
9. Use debugpy in the core extension instead of ptvsd.
   (#10184)
10. Add telemetry for imports in notebooks.
    (#10209)
11. Update data science component to use debugpy.
    (#10211)
12. Use new MacOS VM in Pipelines.
    (#10288)
13. Split the windows PR tests into two sections so they do not time out.
    (#10293)