This is so filled with #fakenews it's perhaps easier to pick out the one useful criticism, which is having to learn 3 string formatting styles, which is indeed

This is so filled with #fakenews it's perhaps easier to pick out the one useful criticism, which is having to learn 3 string formatting styles, which is indeed a pain. For commercial work it seems wise to stdize on one in a given project/organization. As for the rest, I'll pile on: - python3 is eight years old, not decades, and this matters: Perl4 => 5 was a similarly long transition and for similar reasons. - python3 is obviously Turing complete. Moreover, I'm not sure its lack of whatever the author thinks TC means, actually matters in the question of whether python3 is worth choosing to learn or build systems. - very few languages can be used "side by side" with each other, and I want aware that this was a common use case. If you have python2 code, there are various ways to invoke it without porting, eg subprocess, web service, etc. Obviously, there's various costs eg performance, system mgmt, debugging, etc. - high probability of failure? Strange but I see the exact opposite, with more and more projects more supporting ONLY python3 ie the tide is turning. Anyway, this is scaremongering and #fakenews so YMMV.