On Thursday, September 21, 2017 at 10:12:25 AM UTC-5, Steve D'Aprano wrote: > [...] > And remember that the Python core developers feel your pain > too. They had to migrate a large code base (the Python std > library) from 2 to 3. They had to write the 2to3 > translator. And they have to maintain two independent code > bases, 2 and 3, in parallel, for a decade or more, and put > up with thousands of internet haters going on and on and on > and on about "Python 3 is destroying Python", year after > year after year. I think it's grossly unfair to label those who's lives have been up-ended by the backwards incompatible changes of Python3 as "haters". Most of these people did not want or even need these changes, and they were just happy to be using a language that was intuitive, productive and did not distract them from solving whatever problem happened to need solving at the moment.
Sure, i think all the changes are great (except for type hints, and i'm withholding my judgment of async for the time being...), and had these changes been a part of the original language, Python would have been a better language for it. Personally, i will not decide whether or not to migrate until Python4, and i suspect i am not alone in this regard. And if Py-dev feels betrayed by the community, well, they should understand that we feel betrayed as well. We have invested decades into this language, and to have our investments ruined by changes we did not ask for is devastating. So, can we overcome this devastation? Possibly. But Py-dev needs to understand that breaking backwards compatibility is no small thing, and they also need to understand that alienating the Python2.X'ers is not helping the migration to Python3. I assure you, this community cannot endure another one of these "engineered natural disasters". One of the reasons we've been so loyal to Python, is that, at least historically, we had faith in GvR's judgments. And while his fabled time machine has saved our bums on more than one occasion, these days, some of us are not so sure of our capitan's navigational abilities, and are feeling a bit like we are the unlucky passengers on the Titanic. Speaking personally here, i know i could have overcome the backwards incompatible thing _eventually_, but Python3, followed by the introduction of type-hints has been such a earth shattering 1-2-punch, that i'm really struggling to stay on my feet at this point. I mean, i guess i could delude myself and imagine that type- hints don't exist, but that's about as realistic as ignoring a giant dent in the side of my beautiful car. Sure, the car drives just fine, but the ugly dent sours the experience. :-( -- https://mail.python.org/mailman/listinfo/python-list
