The Case Against Python 3

Yea I skimmed the rest. This makes me ashamed that I even bought the guys book (Learn Python the Hard Way). His attitude is what is going to kill python, not python 3. I moved from Perl 5 to Python 3 hoping to flee the internal dev fighting and attitudes, yet here I am again.

I'll say this, at Pycon 2016 I attended some dev sprints/hackathons with the python 3 developers. Python 2 is no where on their radar, its dead. They have moved on. There will be no additional compatibility layers or any of that. All of the mainline libraries that everyone uses (django, sql alchemy, etc ) have moved on to python 3. If you haven't, you should to.

lol, winning.

This has to be the ne plus ultra of Python scaremongering. It describes some sort of bizzaro world of a doomed language which it populates with Machiavellian Python maintainers, brainwashed developers, and a small group of heroic holdouts who see resisting Python 3 as a moral imperative. Apparently having too many string formatting options is a moral issue (three, to be specific).

As someone who writes a _lot_ of Python code (mostly in 3 but with occasional switches to 2), who maintains several libraries which work in both 2 and 3, and as someone who uses a wide range of libraries in stats, machine learning, networking, web development, etc. in my python 3 work...this piece seems totally disconnected from reality. The problems it describes are relentlessly overblown when they're not simply manufactured from whole cloth.

Put another way: the differences between 2 and 3 are not great, the vast majority of libraries you'd want to use (i.e. are actively maintained or of at least good quality) work in 3, and while I don't doubt there's great piles of Python 2 code moldering away big "enterprise" apps and those sorts of places, it's ever been thus in that space, no matter what the language, and doesn't pose any sort of existential threat to Python.

(edit: I used to recommend "Learn Python The Hard Way" to newcomers, and have just kept reflexively doing that over the years because I wasn't aware of a better resource. But if this article has accomplished one thing, it's that it's spurred me to look for a replacement)

I attended a few newbie classes at pycon2016. A few of them suggested Learn Python The Hard Way, to which I pointed out how the author states to never try installing python 3 on the first page.

>C++ is an example of the serious negative consequences of language change.

From the outside looking in (as a non-C++ developer), for me the bigger problem is honestly that they left in the warts rather than just paving around them. Those who have followed C++ for its entire life cycle have probably been able to mostly keep up with what language features were mistakes and what the best way to do things is, but I can't imagine myself ever picking up C++ now and hoping to have any clue at all what to avoid without spending years learning the hard way (no pun intended).

As a very basic example from C# (which I do use professionally), the fact that untyped collections still exist (and things like IEnumerable with explicit casts instead of IEnumerable) is just pointless cruft that we'd be better off without. As far as I'm concerned, no code should ever depend on them, and any code that does should be forcibly broken to force people to fix it. We treat security flaws seriously by forcing people to fix things, why should we not do the same with features that are essentially bug-magnets?