"Problems" re-appear and cannot be hidden after closing a file while it is being linted

Environment data

• VS Code version: 1.46.1
• Extension version (available under the Extensions sidebar): 2020.6.90262
• OS and version: Win 10 2004, Remote SSH into Linux
• Python version (& distribution if applicable, e.g. Anaconda): 3.8.3
• Type of virtual environment used (N/A | venv | virtualenv | conda | …): none
• Relevant/affected Python packages and their versions: e.g., pylint==2.5.3
• Relevant/affected Python-related VS Code extensions and their versions: none, I believe
• Value of the python.languageServer setting: default (Jedi)

Expected behaviour

I think I observed that “Problems” are removed/hidden when a file is closed.

Actual behaviour

However, when I close a file while linting is running (especially if that takes a few seconds), the problems re-appear after the file is closed and cannot be hidden. If “Show Active File Only” is checked, then these problems are not actually shown, but still counted, compare this screenshot:

Steps to reproduce:

1. Create and save a long-linting .py file (the one below works great, takes many seconds)
2. Make a change and save the file
3. Wait for linting to start (you’ll notice the problems in the list disappear), then quickly close the file

I’d expect VS Code to ensure the expected behavior, but it seems to be up to the extenions to ensure this. https://github.com/microsoft/vscode/issues/100646

import humanize
import jstyleson
import matplotlib
import natsort
import nose
import numpy
import ordered_set
import pandas as pd
import psutil
import PyPDF2
import raster_geometry
import scipy
import seaborn
import SharedArray
import silence_tensorflow
import tensorboard
import tensorflow
import tensorflow_addons
import textract
import tf_explain
import tifffile
import toolz

import colour_science
import ExifRead
import Keras
import keras_vis
import opencv_contrib_python
import pynrrd
import python_pptx
import scikit_image
import scikit_learn
import scikit_plot
import Werkzeug


def hello():
    """
    Hello.
    """

    test_pdf = pd.DataFrame([[1, 2, 3]])

[bersbersbers]

The behavior shown in the gif is certainly similar. Certainly the end result is the same: problems continue to be shown for a closed file.

My examples are a bit different in that problems disappear when the linter starts, then re-appear when it’s done. So in that case, it’s even more clear to see what is going on.

Your gif might also show a different, more general problem removing problems for a closing file irrespective of the linter running simultaneously. But assuming that the TS linter implementation is just a little more advanced and just does not hide the previous problems while the linter is running, I’d agree that your gif is compatible with what I describe.

[karthiknadig]

If you run into issue please file a bug on the pylint extension.