Valentin V. Bartenev

Could you please replace your `wsgi_flask.py` with this simple code, that just outputs Python version and test: ```python import platform def application(environ, start_response): response_body = platform.python_version() start_response('200 OK', []) return...

@elder-easy-care Could you also show the output of `printenv PYTHONPATH` command? In this case: ``` # source /git/eec/APPLICATION/www/flask/venv/bin/activate # printenv PYTHONPATH ``` and in this: ``` # deactivate # printenv...

Also please show: ``` # cat /git/eec/APPLICATION/www/flask/venv/pyvenv.cfg ```

It just looks like a simple option, because you're looking at very narrowed use case that barely will work for many of our users. Just a few points those immediately...

Note that our official docker images already provide a pretty handy way of initial configuration, where you can add bootstrap scripts in order to download all the needed files: https://unit.nginx.org/installation/#initial-configuration

@lcts Addressing of such socket name isn't something supported right now. There's a plan to make it possible via URI-escaping, e.g.: `/config/listeners/unix:%2Fpath%2Fto%2Fsocket/`. Sorry, still no updates on automatic deletion. Frankly...

Any ideas on how such configuration may look like? Or maybe we just need to allow configuring access log in the router objects somehow?

@RiderALT If your reasons are related to some functionality missing in Unit or something is better in some other product, then it would be nice if you can share these...

Yes, it's planned.

We definitely don't want additional syscalls on every request. While sometimes it may be handy, but the overhead isn't worth it. I'm sure that most users don't really need it...