8 ms·

The reason, I still did not port to Python 3: (and yes, Unicode in Py2 is a mess ...) They just broke to many things (unnecessarily!) internally. Particularly

The reason, I still did not port to Python 3: (and yes, Unicode in Py2 is a mess ...) They just broke to many things (unnecessarily!) internally. Particularly they changed many C APIs for enhancement modules, so that all of them had to be ported, before they could be used with Python 3. They did not even consider a portability layer ... why not?? Some (not all) of the bad decisions (like the u"..." strings) they did change afterwards, but than it was a little late. So many modules are still not ported to Python 3 -- so the hurdle is a little to high -- for small to nil benefits! So, the problem (from my side) is not Unicode at all ... just the lack of reasonable support from the deciders side. --- Maybe, some time later, when I have to much spare time.

stillsut 11y ago
Agreed. Also this: "So expect Python 4 to not do anything more drastic than to remove maybe deprecated modules from the standard library." But why break all the existing libraries that use those modules, even if there's now "better" ways. In every comparison I've ever seen on performance, robustness, etc python always loses to the other big languages. Except in one area: the availability of user-land submitted packages and extensions. So why break them for a little perf boost?
- PythonicAlpha 11y ago
  Sounds really a bit weird to me. I hope, that they really think hard, which modules really must be removed. I would only consider security reasons to really remove a library module. Or at least mark it deprecated for a long while until it is used only really rarely.