Skip to main content

Return to Answer

added 502 characters in body
Source Link
abarnert
abarnert
  • 367.8k
  • 1
  • 23
  • 29

I think it's worth having a canonical Python 3 answer.

But I think it should be a separate question and answer from the (existing) Python 2 one, rather than trying to merge both into a single question and answer.

The details are very different, and a whole lot simpler in Python 3. Anyone who has a Python 3 question, unless it's actually about porting from Python 2 or writing dual-version code, shouldn't have to slog through all the Python 2 complexities that just make it harder for them to process the answer.

That's already probably the majority of askers today (despite the fact that Unicode errors don't come up as often in Python 3 in the first place), and that will only be more true in the future.

For example: If a Windows user saves a source file with non-ASCII characters in cp1252 via Notepad, they're going to get an error, even in Python 3. But if half the answer is about implicit transcoding between str and unicode in Python 2—which doesn't even have an analog in Python 3—the answer is just going to confuse them. And anyone not using Python 2 in 2018 is probably never going to use it, so learning enough to get past all that confusion isn't even going to buy them anything.

I think it's worth having a canonical Python 3 answer.

But I think it should be a separate question and answer from the (existing) Python 2 one, rather than trying to merge both into a single question and answer.

The details are very different, and a whole lot simpler in Python 3. Anyone who has a Python 3 question, unless it's actually about porting from Python 2 or writing dual-version code, shouldn't have to slog through all the Python 2 complexities that just make it harder for them to process the answer.

That's already probably the majority of askers today (despite the fact that Unicode errors don't come up as often in Python 3 in the first place), and that will only be more true in the future.

I think it's worth having a canonical Python 3 answer.

But I think it should be a separate question and answer from the (existing) Python 2 one, rather than trying to merge both into a single question and answer.

The details are very different, and a whole lot simpler in Python 3. Anyone who has a Python 3 question, unless it's actually about porting from Python 2 or writing dual-version code, shouldn't have to slog through all the Python 2 complexities that just make it harder for them to process the answer.

That's already probably the majority of askers today (despite the fact that Unicode errors don't come up as often in Python 3 in the first place), and that will only be more true in the future.

For example: If a Windows user saves a source file with non-ASCII characters in cp1252 via Notepad, they're going to get an error, even in Python 3. But if half the answer is about implicit transcoding between str and unicode in Python 2—which doesn't even have an analog in Python 3—the answer is just going to confuse them. And anyone not using Python 2 in 2018 is probably never going to use it, so learning enough to get past all that confusion isn't even going to buy them anything.

Source Link
abarnert
abarnert
  • 367.8k
  • 1
  • 23
  • 29

I think it's worth having a canonical Python 3 answer.

But I think it should be a separate question and answer from the (existing) Python 2 one, rather than trying to merge both into a single question and answer.

The details are very different, and a whole lot simpler in Python 3. Anyone who has a Python 3 question, unless it's actually about porting from Python 2 or writing dual-version code, shouldn't have to slog through all the Python 2 complexities that just make it harder for them to process the answer.

That's already probably the majority of askers today (despite the fact that Unicode errors don't come up as often in Python 3 in the first place), and that will only be more true in the future.