I was surprised to recently notice a yellow banner on the website of the very popular Requests library, which urges users to switch to Python 3. That's when I first thought switching may become inevitable. I guess this is being orchestrated behind the scenes now.
Still, I have no plans to switch. The only useful feature in Python 3 to me is more liberal use of unpacking. Unfortunately it comes at the cost of removed tuple parameter unpacking, which I use often, but most users apparently never do. I don't know what's difficult about Unicode in Python 2 either, once you understand the difference between Unicode and UTF-8.
It's unfortunate it ever had to come to this. Makes you wonder what Python would be like today without Py3K. (It's an open question.)
> The only useful feature in Python 3 to me is more liberal use of unpacking.
Unfortunately, this ignores composing software. Your user may use things you don’t. The result: your software won’t get used as a library. That may be fine! More power to you. Just don’t drag anyone else down to 2 with you. :)
Personally, python 2’s print keyword/statement is infuriatingly inconsistent with the rest of the language; the network modules are a mess, organizationally; there’s no async support; the unicode support makes me want to stab my eyes out. I don’t mean to convince you (I’m not very convincing...) just to give an opportunity to hedge your statement with empathy for everyone who did decide to move on. Surely you must have any commentary that doesn’t reduce to “I don’t like change”, right?
Python without py3k is just old software that is end of lifing soon, after all :)
> I don't know what's difficult about Unicode in Python 2 either, once you understand the difference between Unicode and UTF-8.
"once you understand the difference between Unicode and UTF-8" is what's difficult about Unicode in Python 2. I understand it, you might understand it, but I have to interop w/ and work w/ code written by people who do not. I'm not fool-proof either, so I greatly appreciate that the language makes a hard distinction now; doing the right thing by default is the point.
I still use Python 2.7 and all I need with strings is UTF-8 byte string nowadays in Japan.
If you read Japanese, read the following blog to know the current circumstances:
And note that Golang is getting popular in Japan, which uses UTF-8 byte strings solely.
Python 2.7 with byte strings as default has a good chance to evolve into a more elegant language :-)
It has been "orchestrated" ever since PSF announced an official end-of-life date for Python 2.7.
This is just the way things work: support eventually runs out, and if people didn't switch before then for other reasons, that's usually motivation enough to finally do so.
https://docs.python-requests.org
Still, I have no plans to switch. The only useful feature in Python 3 to me is more liberal use of unpacking. Unfortunately it comes at the cost of removed tuple parameter unpacking, which I use often, but most users apparently never do. I don't know what's difficult about Unicode in Python 2 either, once you understand the difference between Unicode and UTF-8.
It's unfortunate it ever had to come to this. Makes you wonder what Python would be like today without Py3K. (It's an open question.)