Python 3 timeline

[hmaarrfk]

Hey @colinoflynn

do you think it is time to drop python2.7? most of the scientific stack is dropping it soon and it is kinda a blocker now for PR #126.

I really have no intention of working hard to ensure that 2.7 works, so I am in favour of dropping it.

Do any of our users need 2.7?

[colinoflynn]

I use 2.7 for a bunch of my work still in ChipWhisperer (based on various other requirements), but I can always just keep an old version around for that. PR #126 didn't fully break 2.7 though it seems, at least on other platforms?

[hmaarrfk]

Yeah, PR #126 didn't "break" any old functionaly that worked with python 2.7, but it definitely doesn't work on Python 2.7.

If 1/2 of the core devs still uses 2.7, then I think it makes sense to keep 2.7 around.

I mostly don't want to create busy work for future contributors. Which we've had a few 😄

[hmaarrfk]

Anyway, leaving this here for reference:
http://python3statement.org/
Many science projects are going to drop 2.7 support this year or next. Should be fun to see how this plays out.

[colinoflynn]

Yeah, it's coming :/ Mostly my problem w/ ChipWhipsperer is PySide, which will likely mean moving to PySide 2 or something else, which ends up being a lot. It may almost be a rewrite when it gets down to it (5 years of changes in architecture add up)... so I wouldn't say hold pico-python back for me! I hardly count as a core dev by commits anymore anyway ;-)

[hmaarrfk]

Take a look at qtpy.

You should be able to write your code in a way that compatible between PySide/PySide2/PyQt4/PyQt5

[colinoflynn]

Will check it out - I think someone was actually working a little on that in one of the branches, but anything to make easier is good. Otherwise I might just open a hotdog stand instead and ignore Python 3 forever.