Allow authors to use different email addresses in different PEPs

[msullivan]

The context here is that Yury Selivanov and I have a forthcoming PEP
and would prefer to use our current employer's email addresses for it,
but don't want to change our email addresses on earlier PEPs we've
written. (Obviously this affects more PEPs for Yury than for me :P)

There's no serious technical impediment to supporting this, so I have.

I've tested this locally.

---

📚 Documentation preview 📚: https://pep-previews--4797.org.readthedocs.build/

[AA-Turner]

I'm not sure we should do this; the name-email pair serves as a unique identifier for authors. We've already had authors with the same surname, and having some way of mapping authors to PEPs in a unique fashion is a useful property.

A

[hugovk]

Previously: #3621

[1st1]

I'm not sure we should do this; the name-email pair serves as a unique identifier for authors. We've already had authors with the same surname, and having some way of mapping authors to PEPs in a unique fashion is a useful property.

Hm, if it's really useful we can map them by some other ID.

But we have a real world use-case here: the PEP we are about to publish is sponsored by the company we work for now, we have to use their email. But we have a bunch of PEPs that were created personally (I have like 8) that I don't want to suddenly show my new work email.

I'd say that having one email <> one name is a very weird restriction and in my current case I'm just not sure what to even do about this.

[AA-Turner]

Is it entirely intractable that you use your current email for the new PEP?

A

[1st1]

Is it entirely intractable that you use your current email for the new PEP?

Well, we've spent months of work on this PEP, it's fair if the employer gets some visibility. It's important. And I think it's in the interest of Python that companies see reason to invest their employees' time into working on PEPs.

[hugovk]

An acknowledgements section would be a more visible way of crediting an employer, and it's good when employers contribute to open source by paying their employees to work on open source.

---

But I think it's okay to allow authors to use different emails in different PEPs, and they can already use no email if they like.

What is the email for?

I'd say it's primarily it's a means to contact the author. True, that might go out of date if they change employer, but we don't require a personal email address, and I'm confident we already have company emails that are now invalid. And likewise for personal emails, I expect some of those are also now invalid.

In fact, if people use multiple emails, this means it's more likely we can find one that works in the future.

Is email used as a unique identifier so we can disambiguate authors? We don't require an email, so that doesn't hold. The build also fails if we had two authors with the same name but different emails, so again, that doesn't hold -- we'd need the author to have a different pen name (PEP name), but that's not reasonable.

And it seems we've not had duplicate author names in ~26 years, so do we even need to disambiguate? If we ever did, and we were confused which author is which (although the pool of potential PEP authors is fairly small), we can look up PRs to disambiguate.

[1st1]

Given @hugovk input -- @AA-Turner can we merge this today? We'd like to publish our PEP.

[hugovk]

Test suggestion added in msullivan#1.

[1st1]

@AA-Turner given that you seem to be away and given that it seems that enough people think this isn't controversial I'll merge this. But please lmk if you think it will have to be reverted, I'll be happy to do that and work with you through any concerns. 🙏