disable default logger if running in debug mode (-D) ...

Debug mode can lead to log sensitive data (like password changes).
Having -D to log exclusively to stderr is a safer option.

fix the urls for modify and delete ...

The id of the user is passed through the querystring in this page.
But the id was not properly escaped to be included as a querystring
parameter leading to weird issues like.

remove the double escaping. ...

Now the escaping is done by in the templates.
We need to remove the previous escaping done by hand in the code.
Otherwise, we end-up with double escaping and funky displaying of
fields.

add warning in documentation for log level 'debug'

version bump + changelog

adding the possibility to log to stdout

changelog and version bump

fix error handling when adding user that already exists

adding (mostly) working configuration example

changelog+version bump

switch to "stable" in setup.py troves

another exception

another exception

another exception

wider exception for <input> attribute "type" has invalid value

another exception for todylib

adding another ignore

another ignore for tidylib

adding a fffew other exception in tidylib

fix typo

html_tidy cleanup

ignoring another nav error

adding exception for <nav> tags in html validation

cleanup in html template + tidylib ...

* few small cleanup in html template (avoid empty tbody, put id between
quotes)
* switch to tidylib to validate the html instead of the previous hack
calling an external service (https://html5.validator.nu/)
* remove the previous validator script
* add exception for tidylib on empty <span> (these are required by
bootstrap)

remove debug print

fix html validator test for python 3

force the groups in flatten roles to be sorted ...

* sorting the groups helps debuggability and also permits testing
that doesn't rely on python ordering (which is different between 2 and
3).

fix test and exception handling in code ...

With python 2 it was possible to do exception[0][...] to recover
details about an exception.
It's no longer authorized with python 3.
Now, we must do something like exception.args or exception.urls.
fortunately this syntax also works with python 2.
So we use it for both.

remove deprecation warning for html escape ...

in python 2, (html) escape is part of the cgi module
in python 3, it's part of the html module

we now do a conditional import depending on the version, and name the
function html_escape.

fix behavior of get_attributes() ...

* make sure it returns an ordered list in both python 2 and python 3