python-dateutil returns "naive" datetime objects

[mlasch]

The Python datetime module from the standard lib supports so called "naive" and "time zone aware" datetime objects: https://docs.python.org/3/library/datetime.html#aware-and-naive-objects

As far as I can see, dateutil returns in some (all?) cases naive datetime objects. Using naive objects is discouraged (is it?) and deprecated since version 3.12 (https://blog.miguelgrinberg.com/post/it-s-time-for-a-change-datetime-utcnow-is-now-deprecated).

Let's say, I want to compare the result from tzstr.transitions() I get a TypeError (Tested with Python3.10):

from datetime import datetime, timezone
import dateutil.tz

now = datetime.now(timezone.utc)  # recommended to replace utcnow()
posix_zone = dateutil.tz.tzstr("CET-1CEST,M3.5.0,M10.5.0/3")
start, end = posix_zone.transitions(2023)
now > start

Also the output of start.timestamp() will depend on the timezone the code is running.

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[27], line 7
      5 posix_zone = dateutil.tz.tzstr("CET-1CEST,M3.5.0,M10.5.0/3")
      6 start, end = posix_zone.transitions(2023)
----> 7 now > start

TypeError: can't compare offset-naive and offset-aware datetimes

This makes sense, as I don't want to compare UTC times with a unspecified time. A workaround for me is to set UTC manually for the returned objects. I have to assume that the native datetime object represents UTC.

start = start.astimezone(timezone.utc)
end = end.astimezone(timezone.utc)

Should dateutil return time zone aware datetime objects, now that it looks like the naive type will be phased out sooner or later?