modernize

[prusnak]

similar to markqvist/LXMF#20

[faragher]

So, my eyes glazed over about halfway through, so I'll have to go back later, but first blush:

Are IOError and OSError identical in all architectures and on all platforms?

Same question with the equivalence of b'foo' and foo.encode("ascii")?

[prusnak]

Are IOError and OSError identical in all architectures and on all platforms?

Yes. It is just an alias. Quoting from the documentation:

Changed in version 3.3: EnvironmentError, IOError, WindowsError, socket.error, select.error and mmap.error have been merged into OSError, and the constructor may return a subclass.

Try this:

>>> IOError
<class 'OSError'>

Same question with the equivalence of b'foo' and foo.encode("ascii")?

Yes

[markqvist]

I really appreciate your willingness to help, but I'm just going to kindly ask if you read the Contribution Guidelines?

Also this one for reference.

[prusnak]

@markqvist I understand the part about down-prioritizing formatting PRs in favor of PRs that provide functional benefits, but since there are no open PRs, I thought it might be worth a shot. If you do not like the changes for whatever reason, you can close the PR and I will carry on. No problem. You are the maintainer after all. :-)

[markqvist]

Looking at your PR, it's seems like you're putting the proper attention into doing everything well and consistently, and the changes are definitely for the better, so in this case, it's mostly that it will probably take a while before I get around to reviewing it all. Because I do have to go over everything meticulously to ensure that nothing bad snuck in, which in this case is going to take a bit of effort ;)

Another thing I'd like you to be aware of is that this also needs to compile for Android via P4A, which has some weird issues with functions inside f-strings in certain cases. So please make sure that it actually works on that platform too. You should be able to confirm it simply by compiling the Sideband APK with your changes in it.

[markqvist]

Oh, and forgot, the "xyz".encode("ascii") are a no-go. Please keep them as b"xyz".

[prusnak]

Oh, and forgot, the "xyz".encode("ascii") are a no-go. Please keep them as b"xyz".

I did the conversions the other way around from encode() to b"".

[markqvist]

Hah! Okay! There you go, didn't even think there were any left. My mistake!

[Erethon]

Is this maybe a chance to settle on a format style and enforce it by checking for non-compliance from now on?

[prusnak]

Is this maybe a chance to settle on a format style and enforce it by checking for non-compliance from now on?

If this is what @markqvist wants, I can create another PR which reformats using black and adds a CI check that checks the format. In either case this should be in another PR, not this one.

[prusnak]

I went through my changes and now I am satisfied with the PR. Removing draft status.