1. 5

  2. 5

    How many more articles do we need that explains that the developers won’t be spending more time on Python 2 since the 1. of January and that their latest product won’t be published before April? I have a feeling that when this wave dies down, there will be a new set of articles explaining that yes, Python 2 is EOL and this is indeed the last release.

    1. 2

      except it is not. RHEL and derivatives will support python 2.7 well into the future

      1. 2

        Ubuntu has eliminated Python 2 and all dependent modules in 20.04.

        RHEL is an extreme case. Yes, Python 2 will be supported for some time to come, but so will a host of other outdated bits and bobs.

        FWIW I agree I don’t really see the point of this article. Just ask the question of anyone on the release team and eliminate your confusion :)

        1. 1

          Ubuntu 18.04 LTS is EOL in April 2023, which is over 3 years away. So there are at least two major distros who will continue supporting python2..

          1. 0

            And HP still supports OpenVMS. My point is that current Linux distros are dropping Python 2.

            Does that mean it will vanish off the face of the earth forever on some magic date? No.

            Does it create downward pressure that has already caused library authors to drop Python 2 compat? Yes.

            Does it mean that Enterprise will SOON begin to feel said downward pressure and likely already has started thinking about a migration path? Also yes.

            So, we can pick nits all day long, but ultimately this transition is a thing and it’s in motion. The vanguard has long since moved, and yes now the long tail needs to catch up. It will. Nothing new under the sun here, the same thing has happened with every piece of software ever.

      2. 1

        This is actually the first I’ve heard of it. It does seem weird that support has already ended but they’re still releasing one more version. There has been a pretty active discussion around this release being unnecessarily confusing (https://discuss.python.org/t/petition-abandon-plans-to-ship-a-2-7-18-in-april/2946)

        It’s a similar situation to Microsoft supporting Windows XP long past it’s original EOL date, which made people and companies more comfortable with not migrating and ultimately made the world slightly worse from a security standpoint.

        The migration to Python 3 has been ongoing for what, 10 years now? And we’ve finally gotten the world sold on a real EOL for Python 2. And then four months later, another update drops for Python 2, and the naysayers have the opportunity to say “See? We don’t need to migrate!”

        If they really wanted the EOL date to coincide with Pycon 2020, that’s the date that should have been communicated. Releasing a version four months after you say you’re done supporting it is just asking for trouble in what has already been 10 years of trouble.